CN112633269A - Logo recognition method and system - Google Patents
Logo recognition method and system Download PDFInfo
- Publication number
- CN112633269A CN112633269A CN202011578226.1A CN202011578226A CN112633269A CN 112633269 A CN112633269 A CN 112633269A CN 202011578226 A CN202011578226 A CN 202011578226A CN 112633269 A CN112633269 A CN 112633269A
- Authority
- CN
- China
- Prior art keywords
- logo
- source domain
- data
- meta
- domain model
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000005070 sampling Methods 0.000 claims abstract description 49
- 230000006870 function Effects 0.000 claims description 32
- 238000013507 mapping Methods 0.000 claims description 24
- 238000012360 testing method Methods 0.000 claims description 22
- 238000012549 training Methods 0.000 claims description 20
- 230000036961 partial effect Effects 0.000 claims description 2
- 230000008569 process Effects 0.000 description 13
- 238000013461 design Methods 0.000 description 12
- 238000004891 communication Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 6
- 230000000007 visual effect Effects 0.000 description 6
- 238000013135 deep learning Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 239000003550 marker Substances 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000012512 characterization method Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 3
- 238000013136 deep learning model Methods 0.000 description 3
- 230000003190 augmentative effect Effects 0.000 description 2
- 230000003796 beauty Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000000670 limiting effect Effects 0.000 description 2
- 238000013508 migration Methods 0.000 description 2
- 230000005012 migration Effects 0.000 description 2
- 238000003909 pattern recognition Methods 0.000 description 2
- 230000000638 stimulation Effects 0.000 description 2
- 238000010200 validation analysis Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000003466 anti-cipated effect Effects 0.000 description 1
- 238000012850 discrimination method Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000002427 irreversible effect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229910052754 neon Inorganic materials 0.000 description 1
- GKAOGPIIYCISHV-UHFFFAOYSA-N neon atom Chemical compound [Ne] GKAOGPIIYCISHV-UHFFFAOYSA-N 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000002829 reductive effect Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000013526 transfer learning Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Computing Systems (AREA)
- Multimedia (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Image Analysis (AREA)
Abstract
The application discloses a Logo identification method and system. The Logo identification method comprises the following steps: selecting: selecting a target domain and a source domain, and grouping the source domains; iteration step: sampling the small-quantity data of the source domain according to a sampling strategy, calculating a loss function of a source domain model, and updating the source domain model by using the sum back propagation of the loss function to obtain a trained source domain model; an acquisition step: and acquiring the category of the unknown LOGO by using the trained source domain model in the target domain. The invention provides a Logo recognition method and a Logo recognition system, wherein the cross-domain problem of Logo is solved through meta-learning, the higher generalization ratio of a model can adapt to a new unknown domain, the requirement on the data volume of the new domain is not high, and the accuracy of Logo recognition of the new domain is improved.
Description
Technical Field
The application relates to the technical field of Logo recognition, in particular to a Logo recognition method and system.
Background
The important premise that general deep learning well performs on a single task is that the deep learning has massive data, better data distribution can drive the deep learning to learn better experiences, and the dependence degree of the general deep learning on the data is larger. Meta-learning is not optimized for a single task, and how to learn, and by capturing the similarity of different tasks, the meta-learning is quickly adapted to a new task, so that the meta-learning has higher generalization and can solve the problem of learning with few samples. When the domain is changed, the performance of the existing deep learning model is often reduced remarkably because the data distribution of different domains is often different, and the general deep learning model is only optimized for the data of the existing domain and is difficult to adapt to the new domain. In the prior art, when a deep learning model faces a new domain, a large amount of data is used for retraining, the requirement on the amount of the data of the new domain is high, and the generalization of the model is low due to the fact that the data in a single domain is only optimized.
Therefore, aiming at the current situation, the invention provides a Logo recognition method and a Logo recognition system, and the cross-domain problem of Logo is solved through meta-learning, so that the higher generalization of the model can adapt to a new unknown domain, the requirement on the data volume of the new domain is not high, and the accuracy of Logo recognition of the new domain is improved.
Disclosure of Invention
The embodiment of the application provides a Logo identification method and system, and aims to at least solve the problem of subjective factor influence in the related technology.
The invention provides a Logo identification method, which comprises the following steps:
selecting: selecting a target domain and a source domain, and grouping the source domains;
iteration step: sampling the small-quantity data of the source domain according to a sampling strategy, calculating a loss function of a source domain model, and updating the source domain model by using the sum back propagation of the loss function to obtain a trained source domain model;
an acquisition step: and acquiring the category of the unknown LOGO by using the trained source domain model in the target domain.
In the Logo identification method, the selecting step includes selecting logos of partial industries, selecting a target domain and a source domain from the logos, and grouping the source domains.
In the Logo identification method, the sampling strategy comprises a meta-training sampling strategy and a meta-testing sampling strategy.
In the aforementioned Logo identification method, the iteration step includes, in each iteration, sampling the small amount of data according to the meta-training sampling strategy and the meta-testing sampling strategy, calculating parameters of a new source domain model in each iteration, calculating the loss function according to the parameters, and updating the source domain model by using the sum of the loss functions to perform back propagation, so as to obtain the trained source domain model.
In the Logo recognition method, the obtaining step includes reasoning data in a library by using the trained source domain model to obtain feature mapping of the data in the library, reasoning the data of the unknown Logo on the target domain by using the trained source domain model, performing feature mapping on the data, and comparing the feature mapping of the data in the library with the feature mapping of the data to obtain the category of the unknown Logo.
The invention provides a Logo recognition system, which is characterized by being applicable to the Logo recognition method, and the Logo recognition system comprises:
a selecting unit: selecting a target domain and a source domain, and grouping the source domains;
an iteration unit: sampling the small-quantity data of the source domain according to a sampling strategy, calculating a loss function of a source domain model, and updating the source domain model by using the sum back propagation of the loss function to obtain a trained source domain model;
an acquisition unit: and acquiring the category of the unknown LOGO by using the trained source domain model in the target domain.
In the Logo identification system, the selecting unit selects a Logo of a part of industries, selects a target domain and a source domain from the Logo, and groups the source domains.
In the Logo identification system, the sampling strategy comprises a meta-training sampling strategy and a meta-testing sampling strategy.
In the Logo identification system, the iteration unit samples the small amount of data according to the meta-training sampling strategy and the meta-testing sampling strategy in each iteration, calculates the parameters of each new round of source domain model, calculates the loss function according to the parameters, and updates the source domain model by using the sum of the loss functions to perform back propagation, so as to obtain the trained source domain model.
In the Logo recognition system, the obtaining unit infers data in a library by using the trained source domain model to obtain feature mapping of the data in the library, infers the data of the unknown Logo by using the trained source domain model in the target domain, and compares the feature mapping of the data in the library with the feature mapping of the data to obtain the category of the unknown Logo after performing the feature mapping on the data.
Compared with the related technology, the invention provides a Logo recognition method and a Logo recognition system, and the cross-domain problem of Logo is solved through meta-learning, so that the higher generalization of the model can adapt to a new unknown domain, the requirement on the data volume of the new domain is not high, and the accuracy of Logo recognition of the new domain is improved.
The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below to provide a more thorough understanding of the application.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
FIG. 1 is a flow diagram of a Logo recognition method according to an embodiment of the present application;
FIG. 2 is a source domain packet display diagram according to an embodiment of the present application;
FIG. 3 is a schematic diagram of the Logo recognition system of the present invention;
fig. 4 is a frame diagram of an electronic device according to an embodiment of the present application.
Wherein the reference numerals are:
a selecting unit: 51;
an iteration unit: 52;
an acquisition unit: 53;
81: a processor;
82: a memory;
83: a communication interface;
80: a bus.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments provided in the present application without any inventive step are within the scope of protection of the present application.
It is obvious that the drawings in the following description are only examples or embodiments of the present application, and that it is also possible for a person skilled in the art to apply the present application to other similar contexts on the basis of these drawings without inventive effort. Moreover, it should be appreciated that in the development of any such actual implementation, as in any engineering or design project, numerous implementation-specific decisions must be made to achieve the developers' specific goals, such as compliance with system-related and business-related constraints, which may vary from one implementation to another.
Reference in the specification 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 specification. 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. Those of ordinary skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments without conflict.
Unless defined otherwise, technical or scientific terms referred to herein shall have the ordinary meaning as understood by those of ordinary skill in the art to which this application belongs. Reference to "a," "an," "the," and similar words throughout this application are not to be construed as limiting in number, and may refer to the singular or the plural. The present application is directed to the use of the terms "including," "comprising," "having," and any variations thereof, which are intended to cover non-exclusive inclusions; for example, a process, method, system, article, or apparatus that comprises a list of steps or modules (elements) is not limited to the listed steps or elements, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus. Reference to "connected," "coupled," and the like in this application is not intended to be limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The term "plurality" as referred to herein means two or more. "and/or" describes an association relationship of associated objects, meaning that three relationships may exist, for example, "A and/or B" may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship. Reference herein to the terms "first," "second," "third," and the like, are merely to distinguish similar objects and do not denote a particular ordering for the objects.
The invention is based on cross-domain Logo recognition of meta-learning, and is briefly introduced below.
Meta-learning (meta-learning) is a mode of machine learning, also called learning how (learning to learning). The main goal of meta learning is to design additional learners so that the machine learning system can automatically learn the objective function from the additional learners. The design of the extra learner is the core of meta-learning, and how to learn the meta-knowledge is also involved. The method is different from the traditional machine learning mode, namely, a task is given, the meta learning is performed on the task, and the meta learning emphasizes that the meta knowledge is abstracted and extracted from a plurality of similar tasks and then applied to a new task. It can be seen that given a task (data set), we can always learn an optimal function using the above process. This process is very versatile. It is anticipated that our machine learning process is very challenging if the number of tasks is very large, or the learning process is very slow. Therefore, it is natural that: how to make maximum use of previously learned tasks to assist in learning new tasks? Transfer learning is one of the effective ideas. Briefly, the migration learning emphasizes that a learned source task exists, the learned source task is directly applied to a target task, and the learning target is achieved through fine adjustment on the target task. This has proven to be an effective way of learning. Meta-learning (also called leaninggto learning) is a very effective learning model. Similar to the goal of migration learning, meta learning also emphasizes learning experience from related tasks to aid in learning of new tasks. The difference is that meta-learning is a more general model, which is centered on the characterization and acquisition of "meta-knowledge". It is understood that this meta-knowledge is a general knowledge that is available over a large class of tasks and is available through some form of learning. It has very powerful characterization capabilities on such tasks and can therefore be generalized to more tasks. To obtain meta-knowledge, in general, meta-learning assumes that we can obtain some tasks, which are sampled from the task distribution. We assume that one can sample from this task distribution a source task, denoted as a training set and a validation set, where the two items represent the training set and validation set, respectively, on one task. Generally, in meta-learning, they are also referred to as support set (support set) and query set (query set). We call the process of learning meta-knowledge the meta-train process, which can be expressed as: which represent the parameters of the meta-knowledge learning process. To verify the effect of meta-knowledge, we define a meta-test procedure: a task is sampled from the task distribution to form meta-test data, denoted as. Thus, during the meta-test process, we can apply the learned meta-knowledge to the meta-test data to train our true task model: it is noted that in the above formula, we are training the parameters of each task adaptively, which completes the generalization process. The basic problems of meta-learning can be divided into three major categories: representation of meta-knowledge (meta-representation). How meta-knowledge should be characterized, this answers the most important question of meta-learning, i.e. the question of what to learn. Meta-learner (meta-optimizer). After the meta-knowledge is characterized, how to select a learning algorithm for optimization, namely, how to learn in the meta-learning is answered. Meta-object (meta-object). With meta-knowledge characterization and learning methods, towards what targets we should learn? This answers the question in meta-learning why this is the case.
The logo is foreign language abbreviation of logo or trademark, is abbreviation of LOGOtype, is a small visual design for identifying identity, plays a role in identifying and popularizing logo owning companies, and can enable consumers to remember company main bodies and brand culture through vivid logos. Logo in the network is mainly a graphic mark used by each website to link with other websites, and represents a website or a plate of the website. LOGO is a variation of Greek LOGO, a modern economic product that differs from ancient stamps. Modern signs bear intangible assets of enterprises and are media for enterprise comprehensive information transfer. The mark is the most important part of the CIS strategy of an enterprise, and is the most widely applied and most frequently occurring element and the most critical element in the process of transmitting the image of the enterprise. The strong overall strength, perfect management mechanism, and high-quality products and services of the enterprise are all covered in the mark, and are deeply remained in the audience center through continuous stimulation and repeated depiction. The concept of the modern logo is more perfect and mature, and the popularization and application of the logo establish a perfect system. With the coming of the digital era and the rapid development of network culture, the traditional information transmission mode and reading mode are challenged unprecedentedly. The conceptual standard of efficiency and time is redefined, and in this case, the style of logo is also shown to be personalized and diversified. For the marker creation and designer, tens of times more information is expressed than before by a compact marker symbol. Through the coexistence of typical logo and design with foreward and exploration tendency, the design tolerance is expanded. The social and economic measuring standard is no longer just the number of commodities, the performance is good or bad, the variety is available, and the accuracy and speed of concept transmission become the key of new measuring standard and winning. It can be said that the era provides an unprecedented practical space for the creation of signs. Based on this, the comprehensive consideration of uniqueness, identifiability, rationality, sensibility, personality, commonality and the like of the mark becomes an effective path for the designers to pursue success. Summarizing the current social and market situation, logo can be roughly summarized in the following several trends. Various marks occupy the visual market of the user in a wide market space, and attract the user. Therefore, how to jump out of the marks is a new requirement for easy identification, easy memory and individuality. Personalization includes personalization of consumer market demand and personalization from designers. Different consumers have different aesthetic orientations, different commodity feelings, and different designers have different originality and performance. Thus, on a diverse platform, personalization is becoming a big trend that is irreversible, both for the consumer market and for designers. Since the end of the 19 th century, the design has tended to be mechanized due to the industrial revolution and the influence of the bahaaus design style, with a cold feeling of a large industrial age. With the development of society and diversification of aesthetics and attention to people, humanization becomes an important factor in design. As noted by the famous Industrial designers, and Proteurs of design educators in the United states, "people always design three dimensions-aesthetics, technology, economy, but more importantly, the fourth dimension-humanity! The same is true for "logo", and the logo should be shaped and shaped according to psychological needs and visual preferences. The aspects of color and the like tend to be humanized and have pertinence. The characteristics of the information age make the prior logo different from the prior logo, and besides showing brand or enterprise attributes, the logo also requires richer visual effect, more vivid modeling, images and color elements which are more suitable for consumption psychology and the like. Meanwhile, translation and creation of self-unique design language are carried out by integrating comprehensive information of multiple aspects of an enterprise, so that the mark can not only express the enterprise idea and the enterprise spirit in an image and in a proper way, but also can cooperate with the market to carry out visual stimulation and attraction on consumers to assist publicity and sale. The mark becomes a visual link and a bridge between the information sender and the information receiver, so that whether the analysis of the information content is accurate or not becomes a way for the logo to win. The aspects of color and the like tend to be humanized and have pertinence. Diversification of consciousness forms enables the artistic expression mode of the logo to be diversified day by day, namely, the logo has a two-dimensional plane form and a common semi-three-dimensional relief concave-convex form; there are stereoscopic marks, there are dynamic neon signs too; the method has a realistic mark and an arbitrary mark; there are strict marks and also conceptual marks. With the progress of network technology and the development of electronic commerce, network signs become increasingly popular new sign forms. The identification of the mark is divided into two categories of artificial mark identification and non-artificial mark identification. The role of the artificial marker is to provide drawing information to the virtual object. The advantage of using a manual marker is that the operator can interact with the virtual object in real time through it. However, in some cases the artificial mark may not be usable, such as digital repair of the garden rounding. It utilizes the augmented reality technology to display virtual buildings in the ruins, thus achieving the effect of reproducing the garden. In this case, the markers cannot be placed in the ruins, and can only be realized by a method without artificial markers. The artificial mark recognition includes mark region recognition and mark pattern recognition, which is an important step for realizing augmented reality. The process of the mark identification comprises the steps of carrying out binarization on an image containing an artificial mark, adopting an algorithm of connected domain extraction to realize mark area identification and the like. There are several methods for the different logo pattern recognition problems, such as: a connected domain number discrimination method and a template matching method. The method better identifies the artificial mark and is an important basis for realizing real-time fusion of virtual and real scenes.
The invention provides a Logo recognition method and a Logo recognition system, wherein the cross-domain problem of Logo is solved through meta-learning, the higher generalization ratio of a model can adapt to a new unknown domain, the requirement on the data volume of the new domain is not high, and the accuracy of Logo recognition of the new domain is improved.
The examples of the present application will be described below with Logo as an example.
Example one
The present embodiment provides a Logo recognition method. Referring to fig. 1-2, fig. 1 is a flow chart of a Logo recognition method according to an embodiment of the present application; fig. 2 is a source domain grouping display diagram according to an embodiment of the present application, and as shown in the diagram, the Logo recognition method includes the following steps:
a selecting step S1: selecting a target domain and a source domain, and grouping the source domains;
iteration step S2: sampling the small-quantity data of the source domain according to a sampling strategy, calculating a loss function of a source domain model, and updating the source domain model by using the sum back propagation of the loss function to obtain a trained source domain model;
acquisition step S3: and acquiring the category of the unknown LOGO by using the trained source domain model in the target domain.
In an embodiment, the selecting step S1 includes selecting a Logo of a part of industries, selecting a target domain and a source domain from the Logo, and grouping the source domains.
In specific implementation, LOGO data of part of industries are selected, wherein the industries comprise: beauty, automotive, food, electronic, sports, apparel, luxury. The meta-learning scheme is now explained by taking beauty, automobiles, food, electronics, sports, apparel as a source domain and luxury as a target domain as examples, and some other industry can be used as a target domain and the rest as a source domain for model training and testing. The industries of the source domain may be divided into 6 groups, as shown in FIG. 2, each group being used by K-1 industries for meta-training and 1 other industry for meta-testing. Group 1 is denoted D1, group D2 is denoted D2, and so on.
In an embodiment, the sampling strategy includes a meta-training sampling strategy and a meta-testing sampling strategy.
In specific implementation, the meta-training sampling strategy is: and M industries are selected from K-1 industries, N samples are selected from each category, and the small data size is (K-1) M N. The meta-test sampling strategy is: from 1 industry, each industry takes M, each category takes N samples, and the small amount of data is M × N.
In an embodiment, the iteration step S2 includes, in each iteration, sampling the small amount of data according to the meta-training sampling strategy and the meta-testing sampling strategy, calculating parameters of each new round of source domain model, calculating the loss function according to the parameters, and updating the source domain model by using the sum of the loss functions to perform back propagation, so as to obtain the trained source domain model.
In specific implementation, in each iteration, only new model parameters are calculated in the meta-training stage according to the loss function, and the model is not updated by back propagation. The meta-test phase computes the contrast and domain alignment loss function L on the new model parameters (obtained in the meta-training phase)meta-test. One L per packetmeta-testThen, sum is performed on all the packets, Σ Lmeta-testAnd the update model is propagated back using this loss function. The loss function contains two parts: comparison (L)contrastiv e) And domain alignment (L)align) Reducing the inter-class spacing and increasing the inter-class spacing by using a loss function of metric learning; and reducing the influence of the domain on LOGO feature distribution by using a domain alignment loss function, so that the LOGO feature distribution is distributed according to the characteristics of the LOGO.
In an embodiment, the obtaining step S3 includes using the trained source domain model to perform inference on data in a library to obtain a feature mapping of the data in the library, using the trained source domain model to perform inference on the data of the unknown LOGO in the target domain, performing feature mapping on the data, and comparing the feature mapping of the data in the library with the feature mapping of the data to obtain the category of the unknown LOGO. The pseudo code for iterating step S2 is as follows:
in the specific implementation, the authentication service is called to carry out unified authority management on the related information, the authority is abnormal, the related information is directly returned to the user side, the authority is normal, the SQL portrait information is built through the SQL portrait module, the related calculation engines are called through the routing module, and the results are merged and returned to the client after the calculation of different engines is completed.
Therefore, the invention provides a Logo recognition method and a Logo recognition system, wherein the cross-domain problem of Logo is solved through meta-learning, the higher generalization of the model can adapt to a new unknown domain, the requirement on the data volume of the new domain is not high, and the accuracy of Logo recognition of the new domain is improved.
Example two
Referring to fig. 3, fig. 3 is a schematic structural diagram of a Logo recognition system according to the present invention. As shown in fig. 3, the Logo recognition system of the present invention is applicable to the Logo recognition method, and includes:
the selecting unit 51: selecting a target domain and a source domain, and grouping the source domains;
the iteration unit 52: sampling the small-quantity data of the source domain according to a sampling strategy, calculating a loss function of a source domain model, and updating the source domain model by using the sum back propagation of the loss function to obtain a trained source domain model;
the acquisition unit 53: and acquiring the category of the unknown LOGO by using the trained source domain model in the target domain.
In this embodiment, the selecting unit 51 selects a Logo of a part of industries, selects a target domain and a source domain from the Logo, and groups the source domains.
In this embodiment, the sampling strategy includes a meta-training sampling strategy and a meta-testing sampling strategy.
In this embodiment, the iteration unit 52 samples the small amount of data according to the meta-training sampling strategy and the meta-testing sampling strategy in each iteration, calculates parameters of each new round of source domain model, calculates the loss function according to the parameters, and updates the source domain model by using the sum of the loss functions to perform back propagation, so as to obtain the trained source domain model.
In this embodiment, the obtaining unit 53 infers data in a library by using the trained source domain model to obtain a feature mapping of the data in the library, infers the data of the unknown LOGO by using the trained source domain model in the target domain, and compares the feature mapping of the data in the library with the feature mapping of the data to obtain the category of the unknown LOGO after performing the feature mapping of the data.
EXAMPLE III
Referring to fig. 4, this embodiment discloses a specific implementation of an electronic device. The electronic device may include a processor 81 and a memory 82 storing computer program instructions.
Specifically, the processor 81 may include a Central Processing Unit (CPU), or A Specific Integrated Circuit (ASIC), or may be configured to implement one or more Integrated circuits of the embodiments of the present Application.
The memory 82 may be used to store or cache various data files for processing and/or communication use, as well as possible computer program instructions executed by the processor 81.
The processor 81 implements any of the Logo recognition methods in the above embodiments by reading and executing computer program instructions stored in the memory 82.
In some of these embodiments, the electronic device may also include a communication interface 83 and a bus 80. As shown in fig. 4, the processor 81, the memory 82, and the communication interface 83 are connected via the bus 80 to complete communication therebetween.
The communication interface 83 is used for implementing communication between modules, devices, units and/or equipment in the embodiment of the present application. The communication port 83 may also be implemented with other components such as: the data communication is carried out among external equipment, image/data acquisition equipment, a database, external storage, an image/data processing workstation and the like.
The bus 80 includes hardware, software, or both to couple the components of the electronic device to one another. Bus 80 includes, but is not limited to, at least one of the following: data Bus (Data Bus), Address Bus (Address Bus), Control Bus (Control Bus), Expansion Bus (Expansion Bus), and Local Bus (Local Bus). By way of example, and not limitation, Bus 80 may include an Accelerated Graphics Port (AGP) or other Graphics Bus, an Enhanced Industry Standard Architecture (EISA) Bus, a Front-Side Bus (FSB), a Hyper Transport (HT) Interconnect, an ISA (ISA) Bus, an InfiniBand (InfiniBand) Interconnect, a Low Pin Count (LPC) Bus, a memory Bus, a microchannel Architecture (MCA) Bus, a PCI (Peripheral Component Interconnect) Bus, a PCI-Express (PCI-X) Bus, a Serial Advanced Technology Attachment (SATA) Bus, a Video Electronics Bus (audio Electronics Association), abbreviated VLB) bus or other suitable bus or a combination of two or more of these. Bus 80 may include one or more buses, where appropriate. Although specific buses are described and shown in the embodiments of the application, any suitable buses or interconnects are contemplated by the application.
The electronic device may be connected to a Logo recognition system to implement the methods described in connection with fig. 1-2.
The technical features of the embodiments described above may be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the embodiments described above are not described, but should be considered as being within the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A Logo recognition method is characterized by comprising the following steps:
selecting: selecting a target domain and a source domain, and grouping the source domains;
iteration step: sampling the small-quantity data of the source domain according to a sampling strategy, calculating a loss function of a source domain model, and updating the source domain model by using the sum back propagation of the loss function to obtain a trained source domain model;
an acquisition step: and acquiring the category of the unknown LOGO by using the trained source domain model in the target domain.
2. The Logo recognition method according to claim 1, wherein the selecting step comprises selecting logos of partial industries, selecting a target domain and a source domain from the logos, and grouping the source domains.
3. The Logo identification method as claimed in claim 1, wherein the sampling strategy comprises a meta training sampling strategy and a meta test sampling strategy.
4. The Logo recognition method according to claim 3, wherein the iteration step comprises, in each iteration, sampling the small amount of data according to the meta-training sampling strategy and the meta-testing sampling strategy, calculating parameters of each new round of source domain model, calculating the loss function according to the parameters, and updating the source domain model by using the sum back propagation of the loss function to obtain the trained source domain model.
5. The Logo recognition method according to claim 1, wherein the obtaining step includes reasoning data in a library by using the trained source domain model to obtain a feature mapping of the data in the library, reasoning the data of the unknown LOGO by using the trained source domain model in the target domain, performing the feature mapping of the data, and comparing the feature mapping of the data in the library with the feature mapping of the data to obtain the category of the unknown LOGO.
6. A Logo recognition system, adapted for use in the Logo recognition method according to any one of claims 1 to 5, the Logo recognition system comprising:
a selecting unit: selecting a target domain and a source domain, and grouping the source domains;
an iteration unit: sampling the small-quantity data of the source domain according to a sampling strategy, calculating a loss function of a source domain model, and updating the source domain model by using the sum back propagation of the loss function to obtain a trained source domain model;
an acquisition unit: and acquiring the category of the unknown LOGO by using the trained source domain model in the target domain.
7. The Logo recognition system according to claim 6, wherein the selecting unit selects a Logo of a part of industries, selects a target domain and a source domain from the Logo, and groups the source domains.
8. The Logo identification system as claimed in claim 7 wherein the sampling strategy comprises a meta training sampling strategy and a meta test sampling strategy.
9. The Logo recognition system as claimed in claim 8, wherein the iteration unit samples the small amount of data according to the meta-training sampling strategy and the meta-testing sampling strategy in each iteration, calculates parameters of a new source domain model in each iteration, calculates the loss function according to the parameters, and updates the source domain model by using the sum of the loss functions to perform back propagation, so as to obtain the trained source domain model.
10. The Logo recognition system as claimed in claim 9, wherein the obtaining unit infers data in a library by using the trained source domain model, obtains a feature mapping of the data in the library, infers the data of the unknown Logo by using the trained source domain model in the target domain, and compares the feature mapping of the data in the library with the feature mapping of the data after performing the feature mapping of the data to obtain the category of the unknown Logo.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011578226.1A CN112633269B (en) | 2020-12-28 | 2020-12-28 | Logo recognition method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011578226.1A CN112633269B (en) | 2020-12-28 | 2020-12-28 | Logo recognition method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112633269A true CN112633269A (en) | 2021-04-09 |
| CN112633269B CN112633269B (en) | 2024-03-15 |
Family
ID=75326036
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011578226.1A Active CN112633269B (en) | 2020-12-28 | 2020-12-28 | Logo recognition method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112633269B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113569934A (en) * | 2021-07-20 | 2021-10-29 | 上海明略人工智能(集团)有限公司 | LOGO classification model construction method and system, electronic device and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106934392A (en) * | 2017-02-28 | 2017-07-07 | 西交利物浦大学 | Vehicle-logo recognition and attribute forecast method based on multi-task learning convolutional neural networks |
| CN108898218A (en) * | 2018-05-24 | 2018-11-27 | 阿里巴巴集团控股有限公司 | A kind of training method of neural network model, device and computer equipment |
| CN110472667A (en) * | 2019-07-19 | 2019-11-19 | 广东工业大学 | Small object classification method based on deconvolution neural network |
| CN110991506A (en) * | 2019-11-22 | 2020-04-10 | 高新兴科技集团股份有限公司 | Vehicle brand identification method, device, equipment and storage medium |
| CN111721536A (en) * | 2020-07-20 | 2020-09-29 | 哈尔滨理工大学 | Rolling bearing fault diagnosis method for improving model migration strategy |
-
2020
- 2020-12-28 CN CN202011578226.1A patent/CN112633269B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106934392A (en) * | 2017-02-28 | 2017-07-07 | 西交利物浦大学 | Vehicle-logo recognition and attribute forecast method based on multi-task learning convolutional neural networks |
| CN108898218A (en) * | 2018-05-24 | 2018-11-27 | 阿里巴巴集团控股有限公司 | A kind of training method of neural network model, device and computer equipment |
| CN110472667A (en) * | 2019-07-19 | 2019-11-19 | 广东工业大学 | Small object classification method based on deconvolution neural network |
| CN110991506A (en) * | 2019-11-22 | 2020-04-10 | 高新兴科技集团股份有限公司 | Vehicle brand identification method, device, equipment and storage medium |
| CN111721536A (en) * | 2020-07-20 | 2020-09-29 | 哈尔滨理工大学 | Rolling bearing fault diagnosis method for improving model migration strategy |
Non-Patent Citations (2)
| Title |
|---|
| JIANZHU GUO等: "Learning Meta Face Recognition in Unseen Domains", 《HTTPS://ARXIV.ORG/ABS/2003.07733》, pages 1 - 10 * |
| 思座の机器学习笔记: "元学习（Meta-learning）——李宏毅老师教学视频笔记", <HTTPS://ZHUANLAN.ZHIHU.COM/P/108503451>, pages 1 - 7 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113569934A (en) * | 2021-07-20 | 2021-10-29 | 上海明略人工智能(集团)有限公司 | LOGO classification model construction method and system, electronic device and storage medium |
| CN113569934B (en) * | 2021-07-20 | 2024-01-23 | 上海明略人工智能(集团)有限公司 | LOGO classification model construction method, LOGO classification model construction system, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112633269B (en) | 2024-03-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111324774B (en) | Video duplicate removal method and device | |
| CN111046275B (en) | User label determining method and device based on artificial intelligence and storage medium | |
| CN113254549B (en) | Character relation mining model training method, character relation mining method and device | |
| CN109685121A (en) | Training method, image search method, the computer equipment of image encrypting algorithm | |
| CN109992781B (en) | Text feature processing method and device and storage medium | |
| CN108959474B (en) | Entity relation extraction method | |
| CN116601626A (en) | Personal knowledge graph construction method and device and related equipment | |
| CN109460503A (en) | Answer input method, device, storage medium and electronic equipment | |
| Demarty et al. | Predicting interestingness of visual content | |
| CN112633269B (en) | Logo recognition method and system | |
| CN115687670A (en) | Image searching method and device, computer readable storage medium and electronic equipment | |
| CN103399900A (en) | Image recommending method based on location service | |
| Pizzati et al. | Manifest: Manifold deformation for few-shot image translation | |
| Wang et al. | Hierarchical image segmentation ensemble for objectness in RGB-D images | |
| CN112767038B (en) | Poster CTR prediction method and device based on aesthetic characteristics | |
| CN114398909A (en) | Question generation method, device, equipment and storage medium for dialogue training | |
| CN112100493B (en) | Document ordering method, device, equipment and storage medium | |
| CN111259975A (en) | Method and device for generating classifier and method and device for classifying text | |
| CN116541592A (en) | Vector generation method, information recommendation method, device, equipment and medium | |
| CN115115869A (en) | Business image labeling method and device, electronic equipment and computer program product | |
| CN108460131A (en) | A kind of tag along sort processing method and processing device | |
| Chambe et al. | Deep learning for assessing the aesthetics of professional photographs | |
| Qin et al. | Statistical system of cultural heritage tourism information based on image feature extraction technology | |
| CN115861605A (en) | Image data processing method, computer equipment and readable storage medium | |
| CN111553335A (en) | Image generation method and apparatus, 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |