CN113298156A - Neural network training method and device for image gender classification - Google Patents
Neural network training method and device for image gender classification Download PDFInfo
- Publication number
- CN113298156A CN113298156A CN202110589208.1A CN202110589208A CN113298156A CN 113298156 A CN113298156 A CN 113298156A CN 202110589208 A CN202110589208 A CN 202110589208A CN 113298156 A CN113298156 A CN 113298156A
- Authority
- CN
- China
- Prior art keywords
- image
- training
- network model
- gender
- feature extraction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 45
- 238000000605 extraction Methods 0.000 claims abstract description 146
- 230000004927 fusion Effects 0.000 claims abstract description 44
- 238000003062 neural network model Methods 0.000 claims abstract description 35
- 230000006870 function Effects 0.000 claims description 32
- 238000004422 calculation algorithm Methods 0.000 claims description 21
- 230000015654 memory Effects 0.000 claims description 17
- 239000000463 material Substances 0.000 abstract description 6
- 238000013145 classification model Methods 0.000 description 6
- 239000000284 extract Substances 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000002708 enhancing effect Effects 0.000 description 4
- 230000001965 increasing effect Effects 0.000 description 4
- 238000002372 labelling Methods 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 230000001131 transforming effect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 2
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000012512 characterization method Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000001537 neural effect Effects 0.000 description 2
- 230000007935 neutral effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a neural network training method and a device for image gender classification, wherein the method comprises the following steps: inputting the training image set into a feature extraction network model for feature extraction to obtain image contour feature information; inputting the training image set into a first gender classification network model to obtain image gender characteristic information; fusing the image contour characteristic information and the image gender characteristic information to obtain image fusion characteristic information; inputting the image fusion characteristic information into a second gender classification network model for training until convergence to obtain a target neural network model; the target neural network model is used for classifying the gender of the input image. Therefore, the method can improve the gender classification accuracy of the network model obtained by subsequent training, and compared with the conventional gender classification network model training method, the method has the advantages of greatly reducing the complexity of the model, accelerating the convergence rate and lowering the cost of manpower and material resources.
Description
Technical Field
The invention relates to the technical field of neural networks, in particular to a neural network training method and device for image gender classification.
Background
In the existing business model, images are more attractive than characters, and the display and promotion effects are more remarkable. Therefore, the images start to have a function of promoting more goods or services, and in this case, it is important how to correctly classify the genders of the images. The existing image gender classification algorithm adopts a multi-label image data set to perform network training, the algorithm is high in cost and not suitable for a single gender classification task, other computer vision tasks such as image segmentation, human body detection or target key point detection are introduced, the algorithm also needs to consume large time cost and labor cost, and meanwhile, the convergence rate of the model is reduced by multi-task parallel training.
Therefore, the existing image gender classification technology has defects and needs to be solved urgently.
Disclosure of Invention
The invention aims to solve the technical problem of providing a neural network training determination method and device for image gender classification, which can introduce the outline information of an image into gender classification network training of the image, improve the gender classification accuracy of a network model obtained by subsequent training, and simultaneously greatly reduce the complexity of the model, accelerate the convergence rate and reduce the cost of manpower and material resources compared with the conventional gender classification network model training method.
In order to solve the technical problem, a first aspect of the present invention discloses a neural network training method for image gender classification, which includes:
inputting the training image set into a feature extraction network model for feature extraction to obtain image contour feature information;
inputting the training image set into a first gender classification network model to obtain image gender characteristic information;
fusing the image contour characteristic information and the image gender characteristic information to obtain image fusion characteristic information;
inputting the image fusion characteristic information into a second gender classification network model for training until convergence to obtain a target neural network model; the target neural network model is used for classifying the gender of the input image.
As an alternative implementation manner, in the first aspect of the present invention, the image contour feature information is multi-channel multi-level image contour feature information; the feature extraction network model comprises a plurality of feature extraction layers for respectively extracting contour features of different levels and a plurality of corresponding uniform size layers for unifying sizes; the method for inputting the training image set into the feature extraction network model to perform feature extraction so as to obtain image contour feature information comprises the following steps:
inputting a training image set to a plurality of the feature extraction layers to output a plurality of image contour features of different sizes;
inputting the image contour features output by each feature extraction layer to the corresponding dimension unification layer to obtain a plurality of image contour features of the same dimension;
and determining the plurality of image contour features of the same size as image contour feature information.
As an alternative embodiment, the plurality of feature extraction layers are a plurality of convolutional layers cascaded in sequence, wherein an output of each convolutional layer is connected to one of the size unification layers.
As an optional implementation manner, in the first aspect of the present invention, the inputting the image fusion feature information into the second gender classification network model for training until convergence to obtain the target neural network model includes:
inputting the image fusion characteristic information into a second gender classification network model for training;
and continuously updating model parameters of the first gender classification network model and/or the second gender classification network model based on back propagation until the first loss function is converged to obtain a target neural network model.
As an optional implementation manner, in the first aspect of the present invention, the first loss function is a softmax loss of the gender prediction information output by the second gender classification network model and the true gender label of the corresponding training image.
As an alternative implementation, in the first aspect of the present invention, the method further includes:
the training image set is processed using a data enhancement algorithm to obtain a training image set comprising more training images.
As an alternative implementation, in the first aspect of the present invention, the method further includes:
inputting the contour training image set into a feature extraction network training model for training until a second loss function of the feature extraction network training model converges to obtain the feature extraction network model; wherein the feature extraction network training model comprises the feature extraction network model and a single-channel feature convolution layer connected to an output of the feature extraction network model.
The invention discloses a neural network training device for image gender classification in a second aspect, which comprises:
the contour extraction module is used for inputting the training image set into the feature extraction network model for feature extraction so as to obtain image contour feature information;
the gender extraction module is used for inputting the training image set into a first gender classification network model to obtain image gender characteristic information;
the fusion module is used for fusing the image contour characteristic information and the image gender characteristic information to obtain image fusion characteristic information;
the training module is used for inputting the image fusion characteristic information into a second gender classification network model for training until convergence to obtain a target neural network model; the target neural network model is used for classifying the gender of the input image.
As an alternative implementation manner, in the second aspect of the present invention, the image contour feature information is multi-channel multi-level image contour feature information; the feature extraction network model comprises a plurality of feature extraction layers for respectively extracting contour features of different levels and a plurality of corresponding uniform size layers for unifying sizes; the contour extraction module comprises:
the feature extraction unit is used for inputting a training image set to the plurality of feature extraction layers so as to output a plurality of image contour features with different sizes;
the size unifying unit is used for inputting the image contour features output by each feature extraction layer to the corresponding size unifying layer so as to obtain a plurality of image contour features with the same size;
and the fusion unit is used for determining the plurality of image contour features with the same size as the image contour feature information.
As an alternative embodiment, the plurality of feature extraction layers are a plurality of convolutional layers cascaded in sequence, wherein an output of each convolutional layer is connected to one of the size unification layers.
As an optional implementation manner, in the second aspect of the present invention, a specific manner in which the training module inputs the image fusion feature information to a second gender classification network model for training until convergence to obtain a target neural network model includes:
inputting the image fusion characteristic information into a second gender classification network model for training;
and continuously updating model parameters of the first gender classification network model and/or the second gender classification network model based on back propagation until the first loss function is converged to obtain a target neural network model.
As an optional implementation manner, in the second aspect of the present invention, the first loss function is a softmax loss of the gender prediction information output by the second gender classification network model and the true gender label of the corresponding training image.
As an alternative embodiment, in the second aspect of the present invention, the apparatus further comprises:
and the data enhancement module is used for processing the training image set by using a data enhancement algorithm to obtain a training image set comprising more training images.
As an alternative embodiment, in the second aspect of the present invention, the apparatus further comprises:
the feature extraction network training module is used for inputting the contour training image set into a feature extraction network training model for training until a second loss function of the feature extraction network training model converges to obtain the feature extraction network model; wherein the feature extraction network training model comprises the feature extraction network model and a single-channel feature convolution layer connected to an output of the feature extraction network model.
The invention discloses a third aspect of the invention discloses another neural network training device for image gender classification, which comprises:
a memory storing executable program code;
a processor coupled with the memory;
the processor calls the executable program code stored in the memory to execute part or all of the steps of the neural network training method for image gender classification disclosed in the first aspect of the embodiment of the invention.
A fourth aspect of the present invention discloses a computer storage medium, where the computer storage medium stores computer instructions, and when the computer instructions are called, the computer instructions are used to perform part or all of the steps in the neural network training method for image gender classification disclosed in the first aspect of the present invention.
Compared with the prior art, the embodiment of the invention has the following beneficial effects:
in the embodiment of the invention, a training image set is input to a feature extraction network model for feature extraction to obtain image contour feature information; inputting the training image set into a first gender classification network model to obtain image gender characteristic information; fusing the image contour characteristic information and the image gender characteristic information to obtain image fusion characteristic information; inputting the image fusion characteristic information into a second gender classification network model for training until convergence to obtain a target neural network model; the target neural network model is used for classifying the gender of the input image. Therefore, the method can simultaneously extract the contour characteristic information and the gender characteristic information of the training image through the two-way network model, and train the gender classification model by combining the characteristics of the two kinds of information after fusion, so that the contour information of the image can be introduced into the gender classification network training of the image, the gender classification accuracy of the network model obtained by subsequent training is improved, and compared with the conventional gender classification network model training method, the method has the advantages of greatly reducing the complexity of the model, increasing the convergence speed and lowering the cost of manpower and material resources.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a schematic flow chart of a neural network training method for image gender classification according to an embodiment of the present disclosure;
FIG. 2 is a schematic flow chart of another neural network training method for image gender classification according to the embodiment of the present disclosure;
FIG. 3 is a schematic structural diagram of a neural network training apparatus for image gender classification according to an embodiment of the present disclosure;
FIG. 4 is a schematic structural diagram of another neural network training apparatus for image gender classification according to the embodiment of the present disclosure;
fig. 5 is a schematic structural diagram of another neural network training apparatus for image gender classification according to an embodiment of the present disclosure.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The terms "first," "second," and the like in the description and claims of the present invention and in the above-described drawings are used for distinguishing between different objects and not for describing a particular order. Furthermore, the terms "include" and "have," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, article, or article that comprises a list of steps or elements is not limited to only those steps or elements listed, but may alternatively include other steps or elements not listed, or inherent to such process, method, article, or article.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the invention. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.
The invention discloses a neural network training method and a device for image gender classification, which can simultaneously extract contour characteristic information and gender characteristic information of a training image through a two-way network model and train a gender classification model by combining the fused characteristics of the two information, thereby introducing the contour information of the image into the gender classification network training of the image, improving the gender classification accuracy of the network model obtained by subsequent training, and simultaneously greatly reducing the complexity of the model, faster convergence rate and lower manpower and material costs compared with the existing gender classification network model training method. The following are detailed below.
Example one
Referring to fig. 1, fig. 1 is a schematic flowchart illustrating a neural network training method for image gender classification according to an embodiment of the present invention. The method described in fig. 1 is applied to a training device of a neural network model, where the training device may be a corresponding training terminal, training device, or server, and the server may be a local server or a cloud server, and the embodiment of the present invention is not limited. As shown in fig. 1, the neural network training method for image gender classification may include the following operations:
101. and inputting the training image set into a feature extraction network model for feature extraction to obtain image contour feature information.
In an embodiment of the present invention, the training image set may include a plurality of training images that are labeled manually, where the manual labeling is to set a gender label of the training images manually for the purpose of implementing the task of gender classification, where the gender label may be one or more of male, female, neutral, and LGBT, and is not limited herein.
In the embodiment of the present invention, the image contour feature is used to represent the outer edge contour information of the image, which may be a combination of one or more of an overall closed contour feature of the image, an edge contour feature of the image, and a texture feature of the image, and the present invention is not limited thereto.
102. And inputting the training image set into a first gender classification network model to obtain image gender characteristic information.
In the embodiment of the invention, the gender classification network model based on the first gender classification network model is used for extracting high-level semantic features used for representing gender information in a training image. Optionally, the first gender classification network model may be one or a combination of more of convolutional neural classification networks such as EfficientNet, ShuffleNet, ResNet, MobileNet, and the like. Optionally, the first gender classification network model may be a gender classification network model obtained by using an image data set for training in advance, for example, using an ImageNet data set for training, and in an optional embodiment, the first gender classification network model receives a garment image with RGB three channels as an input, and finally extracts high-dimensional gender classification features.
103. And fusing the image contour characteristic information and the image gender characteristic information to obtain image fusion characteristic information.
In the embodiment of the invention, the image contour characteristic information and the image gender characteristic information are fused, and the image contour characteristic information and the image gender characteristic information can be spliced in the dimension of each channel and/or subjected to characteristic fusion through a classification characteristic fusion layer formed by a plurality of convolution layers to obtain the image fusion characteristic information. Alternatively, the classification feature fusion layer may be composed of a plurality of cascaded nxn convolutional layers, for example, two cascaded 3 × 3 convolutional layers.
104. And inputting the image fusion characteristic information into a second gender classification network model for training until convergence, so as to obtain a target neural network model.
In the embodiment of the invention, the target neural network model is used for classifying the gender of the input image. Optionally, the architecture of the target neural network model includes the feature extraction network model, the first gender classification network model, and the second gender classification network model, and the data processing flow is similar to the training step, and those skilled in the art will understand that the training step of the neural network model and the actual prediction step have the same or similar technical details, and will not be described herein again. Optionally, the second gender classification network model may be a fully connected layer for gender classification, and the technical details of the first gender classification network model may also be referred to.
Therefore, the method described by the embodiment of the invention can simultaneously extract the contour characteristic information and the gender characteristic information of the training image through the two-way network model, and train the gender classification model by combining the characteristics of the two kinds of information after fusion, so that the contour information of the image can be introduced into the gender classification network training of the image, the gender classification accuracy of the network model obtained by subsequent training is improved, and simultaneously compared with the conventional gender classification network model training method, the method disclosed by the invention has the advantages of greatly reducing the complexity of the model, increasing the convergence rate and lowering the labor cost and material resources.
In an alternative embodiment, the image contour feature information is multi-channel multi-level image contour feature information, and accordingly, the feature extraction network model includes a plurality of feature extraction layers for respectively extracting contour features of different levels and a corresponding plurality of size unification layers for unifying sizes.
Therefore, by implementing the optional implementation manner, the feature extraction network model can be used for extracting contour features of different levels of a plurality of channels of an image, and the sizes of a plurality of feature maps are unified by a size unification layer to obtain multi-level image contour feature information, so that the contour information of the image can be more accurately characterized, and subsequently, when the contour information is introduced into the training of a target neural network, the accuracy of the target neural network for judging the image gender information based on the contour of the image can be improved.
In another alternative embodiment, in step 101, inputting the training image set to a feature extraction network model for feature extraction to obtain image contour feature information, which may include:
inputting a training image set into a plurality of feature extraction layers to output a plurality of image contour features of different sizes;
inputting the image contour features output by each feature extraction layer to a corresponding dimension unification layer to obtain a plurality of image contour features of the same dimension;
a plurality of image contour features of the same size are determined as image contour feature information.
In an embodiment of the present invention, the plurality of feature extraction layers are a plurality of sequentially cascaded convolutional layers, wherein an output of each convolutional layer is connected to a corresponding uniform-dimension layer. Optionally, the feature extraction layer is a convolutional layer including a residual module, where the residual module is used to enhance feature extraction and backward propagation capability, and may be ResNet, densnet or SENet, and finally each feature extraction layer outputs a feature map with different sizes, so as to obtain a feature map with multiple sizes. The shallow feature map is used for capturing fine texture features of the clothing image, and the deep feature map is used for capturing contour features. In an alternative embodiment, the convolutional layer may be a cascade of one 3 × 3 convolutional layer and one residual block.
Optionally, the unified size layer may be an interpolation module, and optionally, the interpolation module performs unified interpolation on the sizes of the image contour features output by the feature extraction layer to the same size by using an interpolation algorithm. Optionally, the interpolation algorithm may be one or more of bilinear interpolation, nearest neighbor interpolation, and deconvolution layer, which is not limited in the present invention.
Therefore, by implementing the optional implementation mode, multi-level image contour characteristic information can be obtained, the contour information of the image can be more accurately characterized, and subsequently, when the contour information is introduced into the training of the target neural network, the accuracy of judging the image gender information by the target neural network based on the contour of the image can be improved.
In yet another alternative embodiment, in step 104, inputting the image fusion feature information into the second gender classification network model for training until convergence, so as to obtain a target neural network model, including:
inputting the image fusion characteristic information into a second gender classification network model for training;
and continuously updating model parameters of the first gender classification network model and/or the second gender classification network model based on back propagation until the first loss function is converged to obtain the target neural network model.
In the embodiment of the present invention, it is preferable that the model parameters of the first gender classification network model and the second gender classification network model are selectively updated, and the model parameters of the feature extraction network model are not updated, because the feature extraction network model in the model is trained in another way, and its function is only used for extracting the image contour features, and if it is trained, the characterization capability of the subsequently extracted image contour features is affected.
Optionally, the final gender prediction information output by the second gender classification network model may be a predicted gender label corresponding to the training image and a corresponding confidence level. Optionally, the first loss function is softmax loss of the gender prediction information output by the second gender classification network model and the real gender label of the corresponding training image, and model parameters of the first gender classification network model and/or the second gender classification network model are continuously updated by calculating back propagation gradient information obtained by the loss, so that the first loss function is converged.
Therefore, by implementing the optional implementation mode, based on back propagation, model parameters of the first gender classification network model and/or the second gender classification network model are continuously updated until the first loss function converges to obtain the trained neural network model, and model parameters of the feature extraction network model can not be updated in training, so that on one hand, the convergence speed in training is improved, the workload is reduced, and on the other hand, the network model can focus on the accuracy of gender classification rather than the accuracy of contour extraction, so as to obtain a better gender classification prediction effect.
In yet another alternative embodiment, the method further comprises the steps of:
and inputting the contour training image set into the feature extraction network training model for training until a second loss function of the feature extraction network training model is converged to obtain the feature extraction network model.
In the embodiment of the invention, the feature extraction network training model comprises a feature extraction network model and a single-channel feature convolution layer connected to the output of the feature extraction network model, wherein the single-channel feature convolution layer is used for processing multi-channel multi-level image contour feature information output by the feature extraction network model into single-channel image contour feature information.
In the embodiment of the invention, the feature extraction network model is obtained by final training, namely the feature extraction network training model without the single-channel feature convolution layer, which is finally used in the scheme of the invention, the feature extraction network model is used for extracting the multi-channel multi-level contour features of the image, the single-channel feature convolution layer is only used for fusing the multi-channel multi-level contour features during training to facilitate subsequent loss calculation, and the training can be abandoned after the training is finished.
Therefore, by implementing the optional implementation mode, the contour training image set can be input into the feature extraction network training model for training until the second loss function of the feature extraction network training model is converged, so that the feature extraction network model for extracting multi-channel and multi-level contour feature information can be obtained, and the efficiency and the accuracy of an image recognition task can be improved when the image recognition task is subsequently performed according to the contour information.
In yet another alternative embodiment, in the above step, inputting the contour training image set into the feature extraction network training model for training until a second loss function of the feature extraction network training model converges to obtain a trained feature extraction network model, including:
inputting the contour training image set into a plurality of feature extraction layers to output a plurality of image contour features of different sizes;
inputting the image contour features output by each feature extraction layer to a corresponding dimension unification layer to obtain a plurality of image contour features of the same dimension;
fusing a plurality of image contour features of the same size to obtain first image contour feature information;
inputting the first image contour feature information into a single-channel feature convolution layer to obtain single-channel image contour feature information;
and repeating the steps, and updating the model parameters of the feature extraction network model based on the back propagation until the first loss function is converged to obtain the trained feature extraction network model.
In an embodiment of the present invention, optionally, the contour training image set includes a plurality of labeled contour training images, where the labeled contour training images are training images obtained by manually labeling contours in the images. Optionally, the second loss function is a cross entropy loss between the image contour feature information of the single channel and the corresponding contour training image.
Therefore, the optional implementation mode can train the feature extraction network training model until the second loss function of the feature extraction network training model is converged, so that the feature extraction network model for extracting multi-channel and multi-level contour feature information can be obtained, and the efficiency and the accuracy of an image recognition task can be improved when the image recognition task is subsequently performed according to the contour information.
Example two
Referring to fig. 2, fig. 2 is a schematic flowchart illustrating another neural network training method for image gender classification according to an embodiment of the present invention. The method described in fig. 2 is applied to a training device of a neural network model, where the training device may be a corresponding training terminal, training device, or server, and the server may be a local server or a cloud server, and the embodiment of the present invention is not limited. As shown in fig. 2, the neural network training method for image gender classification may include the following operations:
201. the training image set is processed using a data enhancement algorithm to obtain a training image set comprising more training images.
202. And inputting the training image set into a feature extraction network model for feature extraction to obtain image contour feature information.
203. And inputting the training image set into a first gender classification network model to obtain image gender characteristic information.
204. And fusing the image contour characteristic information and the image gender characteristic information to obtain image fusion characteristic information.
205. And inputting the image fusion characteristic information into a second gender classification network model for training until convergence, so as to obtain a target neural network model.
The detailed technical details and technical noun explanation of the above step 202-.
In the embodiment of the present invention, the data enhancement algorithm may be an offline data enhancement algorithm or an online data enhancement algorithm, and it may be a data enhancement method for transforming information such as size, direction, color, resolution, etc. of the training image, for example, it may be one or more of processing operations such as flipping, rotating, clipping, scaling, translating, affine transformation, adding noise, enhancing brightness, enhancing contrast, sharpening, etc.,
therefore, the training image set can be processed by using the data enhancement algorithm to obtain the training image set comprising more training images, so that the amount of the training image data is increased under the condition of reducing the workload, the degree of model training is further improved, and the prediction accuracy of the trained model is improved.
In another alternative embodiment, in step 201, the training image set is processed by using a data enhancement algorithm to obtain a training image set including more training images, including:
the color information of one or more training images in the set of training images is transformed to obtain a set of training images comprising more training images.
Optionally, the manner of transforming the color information may include randomly exchanging color channels, and randomly changing one or a combination of two of the magnitudes of the characteristic values of the specific channels, which is not limited in the present invention. Optionally, the degree of color information transformation here should be smaller than a preset threshold, for example, the number of pictures for exchanging color channels should be smaller than a number threshold, or the difference value for changing the size of the feature value of a specific channel should be smaller than a difference threshold, so as to prevent the color data distribution of the whole garment image from being affected, resulting in the decrease of model accuracy.
Therefore, the alternative embodiment can reduce the possibility that the finally trained gender classification model directly outputs gender categories by means of color information, thereby improving the generalization of the model.
It should be noted that the neural network training method for image gender classification described in the foregoing or following embodiments of the present invention may be specifically applied to the field of gender classification prediction of clothing images, and this field often uses contour information to characterize gender, and the corresponding training image or contour training image in the present invention may be a clothing commodity image, of course, optionally, other fields of commodity images or service images may also be applicable to the method described in the present invention, and the present invention is not particularly limited, but only illustrates the above preferred cases.
EXAMPLE III
Referring to fig. 3, fig. 3 is a schematic structural diagram of a neural network training device for image gender classification according to an embodiment of the present invention. The apparatus described in fig. 3 may be applied to a corresponding training terminal, training device, or server, and the server may be a local server or a cloud server, which is not limited in the embodiment of the present invention. As shown in fig. 3, the apparatus may include:
the contour extraction module 301 is configured to input the training image set to a feature extraction network model for feature extraction, so as to obtain image contour feature information.
In an embodiment of the present invention, the training image set may include a plurality of training images that are labeled manually, where the manual labeling is to set a gender label of the training images manually for the purpose of implementing the task of gender classification, where the gender label may be one or more of male, female, neutral, and LGBT, and is not limited herein.
In the embodiment of the present invention, the image contour feature is used to represent the outer edge contour information of the image, which may be a combination of one or more of an overall closed contour feature of the image, an edge contour feature of the image, and a texture feature of the image, and the present invention is not limited thereto.
A gender extraction module 302, configured to input the training image set to the first gender classification network model to obtain image gender feature information.
In the embodiment of the invention, the gender classification network model based on the first gender classification network model is used for extracting high-level semantic features used for representing gender information in a training image. Optionally, the first gender classification network model may be one or a combination of more of convolutional neural classification networks such as EfficientNet, ShuffleNet, ResNet, MobileNet, and the like. Optionally, the first gender classification network model may be a gender classification network model obtained by using an image data set for training in advance, for example, using an ImageNet data set for training, and in an optional embodiment, the first gender classification network model receives a garment image with RGB three channels as an input, and finally extracts high-dimensional gender classification features.
And the fusion module 303 is configured to fuse the image contour feature information and the image gender feature information to obtain image fusion feature information.
In the embodiment of the invention, the image contour characteristic information and the image gender characteristic information are fused, and the image contour characteristic information and the image gender characteristic information can be spliced in the dimension of each channel and/or subjected to characteristic fusion through a classification characteristic fusion layer formed by a plurality of convolution layers to obtain the image fusion characteristic information. Alternatively, the classification feature fusion layer may be composed of a plurality of cascaded nxn convolutional layers, for example, two cascaded 3 × 3 convolutional layers.
And the training module 304 is configured to input the image fusion feature information to the second gender classification network model for training until convergence, so as to obtain a target neural network model.
In the embodiment of the invention, the target neural network model is used for classifying the gender of the input image. Optionally, the architecture of the target neural network model includes the feature extraction network model, the first gender classification network model, and the second gender classification network model, and the data processing flow is similar to the training step, and those skilled in the art will understand that the training step of the neural network model and the actual prediction step have the same or similar technical details, and will not be described herein again. Optionally, the second gender classification network model may be a fully connected layer for gender classification, and the technical details of the first gender classification network model may also be referred to.
Therefore, the device described by the embodiment of the invention can simultaneously extract the contour characteristic information and the gender characteristic information of the training image through the two-way network model, and train the gender classification model by combining the characteristics of the two kinds of information after fusion, so that the contour information of the image can be introduced into the gender classification network training of the image, the gender classification accuracy of the network model obtained by subsequent training is improved, and compared with the conventional gender classification network model training method, the device disclosed by the embodiment of the invention has the advantages that the complexity of the model is greatly reduced, the convergence speed is higher, and the labor cost and the material resources are lower.
In an alternative embodiment, the image contour feature information is multi-channel multi-level image contour feature information, and accordingly, the feature extraction network model includes a plurality of feature extraction layers for respectively extracting contour features of different levels and a corresponding plurality of size unification layers for unifying sizes.
Therefore, by implementing the optional implementation manner, the feature extraction network model can be used for extracting contour features of different levels of a plurality of channels of an image, and the sizes of a plurality of feature maps are unified by a size unification layer to obtain multi-level image contour feature information, so that the contour information of the image can be more accurately characterized, and subsequently, when the contour information is introduced into the training of a target neural network, the accuracy of the target neural network for judging the image gender information based on the contour of the image can be improved.
In an alternative embodiment, as shown in fig. 4, the contour extraction module 301 comprises:
a feature extraction unit 3011, configured to input a training image set to a plurality of feature extraction layers to output a plurality of image contour features of different sizes;
a size unifying unit 3012, configured to input the image contour features output by each feature extraction layer to a corresponding size unifying layer, so as to obtain a plurality of image contour features of the same size;
and a fusion unit 3013, configured to determine a plurality of image contour features of the same size as the image contour feature information.
In an embodiment of the present invention, the plurality of feature extraction layers are a plurality of sequentially cascaded convolutional layers, wherein an output of each convolutional layer is connected to a corresponding uniform-dimension layer. Optionally, the feature extraction layer is a convolutional layer including a residual module, where the residual module is used to enhance feature extraction and backward propagation capability, and may be ResNet, densnet or SENet, and finally each feature extraction layer outputs a feature map with different sizes, so as to obtain a feature map with multiple sizes. The shallow feature map is used for capturing fine texture features of the clothing image, and the deep feature map is used for capturing contour features. In an alternative embodiment, the convolutional layer may be a cascade of one 3 × 3 convolutional layer and one residual block.
Optionally, the unified size layer may be an interpolation module, and optionally, the interpolation module performs unified interpolation on the sizes of the image contour features output by the feature extraction layer to the same size by using an interpolation algorithm. Optionally, the interpolation algorithm may be one or more of bilinear interpolation, nearest neighbor interpolation, and deconvolution layer, which is not limited in the present invention.
Therefore, by implementing the optional implementation mode, multi-level image contour characteristic information can be obtained, the contour information of the image can be more accurately characterized, and subsequently, when the contour information is introduced into the training of the target neural network, the accuracy of judging the image gender information by the target neural network based on the contour of the image can be improved.
As an optional implementation manner, the specific manner in which the training module 304 inputs the image fusion feature information to the second gender classification network model for training until convergence to obtain the target neural network model includes:
inputting the image fusion characteristic information into a second gender classification network model for training;
and continuously updating model parameters of the first gender classification network model and/or the second gender classification network model based on back propagation until the first loss function is converged to obtain the target neural network model.
In the embodiment of the present invention, it is preferable that the model parameters of the first gender classification network model and the second gender classification network model are selectively updated, and the model parameters of the feature extraction network model are not updated, because the feature extraction network model in the model is trained in another way, and its function is only used for extracting the image contour features, and if it is trained, the characterization capability of the subsequently extracted image contour features is affected.
Optionally, the final gender prediction information output by the second gender classification network model may be a predicted gender label corresponding to the training image and a corresponding confidence level. Optionally, the first loss function is softmax loss of the gender prediction information output by the second gender classification network model and the real gender label of the corresponding training image, and model parameters of the first gender classification network model and/or the second gender classification network model are continuously updated by calculating back propagation gradient information obtained by the loss, so that the first loss function is converged.
Therefore, by implementing the optional implementation mode, based on back propagation, model parameters of the first gender classification network model and/or the second gender classification network model are continuously updated until the first loss function converges to obtain the trained neural network model, and model parameters of the feature extraction network model can not be updated in training, so that on one hand, the convergence speed in training is improved, the workload is reduced, and on the other hand, the network model can focus on the accuracy of gender classification rather than the accuracy of contour extraction, so as to obtain a better gender classification prediction effect.
As an optional implementation, the apparatus further comprises:
and the feature extraction network training module is used for inputting the contour training image set into the feature extraction network training model for training until a second loss function of the feature extraction network training model is converged to obtain the feature extraction network model.
In the embodiment of the invention, the feature extraction network training model comprises a feature extraction network model and a single-channel feature convolution layer connected to the output of the feature extraction network model, wherein the single-channel feature convolution layer is used for processing multi-channel multi-level image contour feature information output by the feature extraction network model into single-channel image contour feature information.
In the embodiment of the invention, the feature extraction network model is obtained by final training, namely the feature extraction network training model without the single-channel feature convolution layer, which is finally used in the scheme of the invention, the feature extraction network model is used for extracting the multi-channel multi-level contour features of the image, the single-channel feature convolution layer is only used for fusing the multi-channel multi-level contour features during training to facilitate subsequent loss calculation, and the training can be abandoned after the training is finished.
Therefore, by implementing the optional implementation mode, the contour training image set can be input into the feature extraction network training model for training until the second loss function of the feature extraction network training model is converged, so that the feature extraction network model for extracting multi-channel and multi-level contour feature information can be obtained, and the efficiency and the accuracy of an image recognition task can be improved when the image recognition task is subsequently performed according to the contour information.
In yet another alternative embodiment, the inputting of the contour training image set into the feature extraction network training model by the feature extraction network training module for training until the first loss function of the feature extraction network training model converges to obtain a specific mode of the trained feature extraction network model includes:
inputting the contour training image set into a plurality of feature extraction layers to output a plurality of image contour features of different sizes;
inputting the image contour features output by each feature extraction layer to a corresponding dimension unification layer to obtain a plurality of image contour features of the same dimension;
fusing a plurality of image contour features of the same size to obtain first image contour feature information;
inputting the first image contour feature information into a single-channel feature convolution layer to obtain single-channel image contour feature information;
and repeating the steps, and updating the model parameters of the feature extraction network model based on the back propagation until the first loss function is converged to obtain the trained feature extraction network model.
In an embodiment of the present invention, optionally, the contour training image set includes a plurality of labeled contour training images, where the labeled contour training images are training images obtained by manually labeling contours in the images. Optionally, the second loss function is a cross entropy loss between the image contour feature information of the single channel and the corresponding contour training image.
Therefore, the optional implementation mode can train the feature extraction network training model until the second loss function of the feature extraction network training model is converged, so that the feature extraction network model for extracting multi-channel and multi-level contour feature information can be obtained, and the efficiency and the accuracy of an image recognition task can be improved when the image recognition task is subsequently performed according to the contour information.
As an optional implementation, the apparatus further comprises:
and the data enhancement module is used for processing the training image set by using a data enhancement algorithm to obtain a training image set comprising more training images.
In the embodiment of the present invention, the data enhancement algorithm may be an offline data enhancement algorithm or an online data enhancement algorithm, and it may be a data enhancement method for transforming information such as size, direction, color, resolution, etc. of the training image, for example, it may be one or more of processing operations such as flipping, rotating, clipping, scaling, translating, affine transformation, adding noise, enhancing brightness, enhancing contrast, sharpening, etc.,
therefore, in the optional embodiment, the training image set can be processed by using the data enhancement algorithm to obtain the training image set including more training images, so that the amount of the training image data is increased under the condition of reducing the workload, the degree of model training is further improved, and the prediction accuracy of the trained model is improved.
As an alternative embodiment, the specific way in which the data enhancement module processes the training image set using the data enhancement algorithm to obtain the training image set including more training images includes:
the color information of one or more training images in the set of training images is transformed to obtain a set of training images comprising more training images.
Optionally, the manner of transforming the color information may include randomly exchanging color channels, and randomly changing one or a combination of two of the magnitudes of the characteristic values of the specific channels, which is not limited in the present invention. Optionally, the degree of color information transformation here should be smaller than a preset threshold, for example, the number of pictures for exchanging color channels should be smaller than a number threshold, or the difference value for changing the size of the feature value of a specific channel should be smaller than a difference threshold, so as to prevent the color data distribution of the whole garment image from being affected, resulting in the decrease of model accuracy.
Therefore, the alternative embodiment can reduce the possibility that the finally trained gender classification model directly outputs gender categories by means of color information, thereby improving the generalization of the model.
Example four
Referring to fig. 5, fig. 5 is a schematic structural diagram of another neural network training device for image gender classification according to an embodiment of the present invention. As shown in fig. 5, the apparatus may include:
a memory 401 storing executable program code;
a processor 402 coupled with the memory 401;
the processor 402 calls the executable program code stored in the memory 401 to execute part or all of the steps of the neural network training method for image gender classification disclosed in the first embodiment or the second embodiment of the present invention.
EXAMPLE five
The embodiment of the invention discloses a computer storage medium, which stores computer instructions, and when the computer instructions are called, the computer instructions are used for executing part or all of the steps of the neural network training method for image gender classification disclosed in the first embodiment or the second embodiment of the invention.
The above-described embodiments of the apparatus are merely illustrative, and the modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, may be located in one place, or may be distributed on a plurality of network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above detailed description of the embodiments, those skilled in the art will clearly understand that the embodiments may be implemented by software plus a necessary general hardware platform, and may also be implemented by hardware. Based on such understanding, the above technical solutions may be embodied in the form of a software product, which may be stored in a computer-readable storage medium, where the storage medium includes a Read-Only Memory (ROM), a Random Access Memory (RAM), a Programmable Read-Only Memory (PROM), an Erasable Programmable Read-Only Memory (EPROM), a One-time Programmable Read-Only Memory (OTPROM), an Electrically Erasable Programmable Read-Only Memory (EEPROM), a Compact Disc-Read-Only Memory (CD-ROM), or other disk memories, CD-ROMs, or other magnetic disks, A tape memory, or any other medium readable by a computer that can be used to carry or store data.
Finally, it should be noted that: the neural network training method and apparatus for image gender classification disclosed in the embodiments of the present invention are only disclosed as preferred embodiments of the present invention, which are only used for illustrating the technical solutions of the present invention, but not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art; the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. A neural network training method for image gender classification, the method comprising:
inputting the training image set into a feature extraction network model for feature extraction to obtain image contour feature information;
inputting the training image set into a first gender classification network model to obtain image gender characteristic information;
fusing the image contour characteristic information and the image gender characteristic information to obtain image fusion characteristic information;
inputting the image fusion characteristic information into a second gender classification network model for training until convergence to obtain a target neural network model; the target neural network model is used for classifying the gender of the input image.
2. The neural network training method for image gender classification as claimed in claim 1, wherein the image contour feature information is multi-channel multi-level image contour feature information; the feature extraction network model comprises a plurality of feature extraction layers for respectively extracting contour features of different levels and a plurality of corresponding uniform size layers for unifying sizes; the method for inputting the training image set into the feature extraction network model to perform feature extraction so as to obtain image contour feature information comprises the following steps:
inputting a training image set to a plurality of the feature extraction layers to output a plurality of image contour features of different sizes;
inputting the image contour features output by each feature extraction layer to the corresponding dimension unification layer to obtain a plurality of image contour features of the same dimension;
and determining the plurality of image contour features of the same size as image contour feature information.
3. The method of claim 2, wherein the plurality of feature extraction layers are a plurality of convolutional layers cascaded in sequence, wherein an output of each convolutional layer is connected to one of the dimensional uniformity layers.
4. The method for training the neural network for image gender classification as claimed in claim 1, wherein the inputting the image fusion feature information to a second gender classification network model for training until convergence to obtain a target neural network model comprises:
inputting the image fusion characteristic information into a second gender classification network model for training;
and continuously updating model parameters of the first gender classification network model and/or the second gender classification network model based on back propagation until the first loss function is converged to obtain a target neural network model.
5. The neural network training method for image gender classification as claimed in claim 4, wherein the first loss function is softmax loss of gender prediction information output by the second gender classification network model and true gender labels of corresponding training images.
6. The neural network training method for image gender classification as claimed in claim 1, further comprising:
the training image set is processed using a data enhancement algorithm to obtain a training image set comprising more training images.
7. The neural network training method for image gender classification as claimed in claim 6, further comprising:
inputting the contour training image set into a feature extraction network training model for training until a second loss function of the feature extraction network training model converges to obtain the feature extraction network model; wherein the feature extraction network training model comprises the feature extraction network model and a single-channel feature convolution layer connected to an output of the feature extraction network model.
8. An apparatus for neural network training for image gender classification, the apparatus comprising:
the contour extraction module is used for inputting the training image set into the feature extraction network model for feature extraction so as to obtain image contour feature information;
the gender extraction module is used for inputting the training image set into a first gender classification network model to obtain image gender characteristic information;
the fusion module is used for fusing the image contour characteristic information and the image gender characteristic information to obtain image fusion characteristic information;
the training module is used for inputting the image fusion characteristic information into a second gender classification network model for training until convergence to obtain a target neural network model; the target neural network model is used for classifying the gender of the input image.
9. An apparatus for neural network training for image gender classification, the apparatus comprising:
a memory storing executable program code;
a processor coupled with the memory;
the processor invokes the executable program code stored in the memory to perform the neural network training method for image gender classification as claimed in any one of claims 1 to 7.
10. A computer storage medium storing computer instructions which, when invoked, perform the neural network training method for image gender classification as recited in any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110589208.1A CN113298156A (en) | 2021-05-28 | 2021-05-28 | Neural network training method and device for image gender classification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110589208.1A CN113298156A (en) | 2021-05-28 | 2021-05-28 | Neural network training method and device for image gender classification |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113298156A true CN113298156A (en) | 2021-08-24 |
Family
ID=77325900
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110589208.1A Pending CN113298156A (en) | 2021-05-28 | 2021-05-28 | Neural network training method and device for image gender classification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113298156A (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105095833A (en) * | 2014-05-08 | 2015-11-25 | 中国科学院声学研究所 | Network constructing method for human face identification, identification method and system |
| CN106295591A (en) * | 2016-08-17 | 2017-01-04 | 乐视控股（北京）有限公司 | Gender identification method based on facial image and device |
| CN106815566A (en) * | 2016-12-29 | 2017-06-09 | 天津中科智能识别产业技术研究院有限公司 | A kind of face retrieval method based on multitask convolutional neural networks |
| CN107301389A (en) * | 2017-06-16 | 2017-10-27 | 广东欧珀移动通信有限公司 | Based on face characteristic identification user's property method for distinguishing, device and terminal |
| CN109035251A (en) * | 2018-06-06 | 2018-12-18 | 杭州电子科技大学 | One kind being based on the decoded image outline detection method of Analysis On Multi-scale Features |
| CN109118487A (en) * | 2018-08-23 | 2019-01-01 | 合肥工业大学 | Bone age assessment method based on non-down sampling contourlet transform and convolutional neural networks |
| CN109271957A (en) * | 2018-09-30 | 2019-01-25 | 厦门市巨龙信息科技有限公司 | Face gender identification method and device |
| CN110110715A (en) * | 2019-04-30 | 2019-08-09 | 北京金山云网络技术有限公司 | Text detection model training method, text filed, content determine method and apparatus |
| CN110502959A (en) * | 2018-05-17 | 2019-11-26 | Oppo广东移动通信有限公司 | Sexual discriminating method, apparatus, storage medium and electronic equipment |
| WO2021087985A1 (en) * | 2019-11-08 | 2021-05-14 | 深圳市欢太科技有限公司 | Model training method and apparatus, storage medium, and electronic device |
-
2021
- 2021-05-28 CN CN202110589208.1A patent/CN113298156A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105095833A (en) * | 2014-05-08 | 2015-11-25 | 中国科学院声学研究所 | Network constructing method for human face identification, identification method and system |
| CN106295591A (en) * | 2016-08-17 | 2017-01-04 | 乐视控股（北京）有限公司 | Gender identification method based on facial image and device |
| CN106815566A (en) * | 2016-12-29 | 2017-06-09 | 天津中科智能识别产业技术研究院有限公司 | A kind of face retrieval method based on multitask convolutional neural networks |
| CN107301389A (en) * | 2017-06-16 | 2017-10-27 | 广东欧珀移动通信有限公司 | Based on face characteristic identification user's property method for distinguishing, device and terminal |
| CN110502959A (en) * | 2018-05-17 | 2019-11-26 | Oppo广东移动通信有限公司 | Sexual discriminating method, apparatus, storage medium and electronic equipment |
| CN109035251A (en) * | 2018-06-06 | 2018-12-18 | 杭州电子科技大学 | One kind being based on the decoded image outline detection method of Analysis On Multi-scale Features |
| CN109118487A (en) * | 2018-08-23 | 2019-01-01 | 合肥工业大学 | Bone age assessment method based on non-down sampling contourlet transform and convolutional neural networks |
| CN109271957A (en) * | 2018-09-30 | 2019-01-25 | 厦门市巨龙信息科技有限公司 | Face gender identification method and device |
| CN110110715A (en) * | 2019-04-30 | 2019-08-09 | 北京金山云网络技术有限公司 | Text detection model training method, text filed, content determine method and apparatus |
| WO2021087985A1 (en) * | 2019-11-08 | 2021-05-14 | 深圳市欢太科技有限公司 | Model training method and apparatus, storage medium, and electronic device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112132156B (en) | Image saliency target detection method and system based on multi-depth feature fusion | |
| CN109002755B (en) | Age estimation model construction method and estimation method based on face image | |
| CN114463586A (en) | Training and image recognition method, device, equipment and medium of image recognition model | |
| CN111914654B (en) | Text layout analysis method, device, equipment and medium | |
| CN108960412B (en) | Image recognition method, device and computer readable storage medium | |
| CN113704531A (en) | Image processing method, image processing device, electronic equipment and computer readable storage medium | |
| CN113591719B (en) | Natural scene arbitrary shape text detection method, device and training method | |
| CN111932577B (en) | Text detection method, electronic device and computer readable medium | |
| CN112926565B (en) | Picture text recognition method, system, equipment and storage medium | |
| CN111767906A (en) | Face detection model training method, face detection device and electronic equipment | |
| CN113822951A (en) | Image processing method, image processing device, electronic equipment and storage medium | |
| CN111680690A (en) | Character recognition method and device | |
| CN110852327A (en) | Image processing method, image processing device, electronic equipment and storage medium | |
| Li et al. | Gated auxiliary edge detection task for road extraction with weight-balanced loss | |
| CN114581710A (en) | Image recognition method, device, equipment, readable storage medium and program product | |
| CN112884758A (en) | Defective insulator sample generation method and system based on style migration method | |
| CN111652181A (en) | Target tracking method and device and electronic equipment | |
| CN114445826A (en) | Visual question answering method and device, electronic equipment and storage medium | |
| CN113298156A (en) | Neural network training method and device for image gender classification | |
| CN116051593A (en) | Clothing image extraction method and device, equipment, medium and product thereof | |
| CN113298092A (en) | Neural network training method and device for extracting multi-level image contour information | |
| CN113240071A (en) | Graph neural network processing method and device, computer equipment and storage medium | |
| CN113392840B (en) | Real-time semantic segmentation method based on multi-scale segmentation fusion | |
| CN117593610B (en) | Image recognition network training and deployment and recognition methods, devices, equipment and media | |
| CN112016554B (en) | Semantic segmentation method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |