WO2018112514A1 - Systèmes et procédés d'apprentissage profond destinés à être utilisés dans la vision artificielle - Google Patents

Systèmes et procédés d'apprentissage profond destinés à être utilisés dans la vision artificielle Download PDF

Info

Publication number
WO2018112514A1
WO2018112514A1 PCT/AU2017/051388 AU2017051388W WO2018112514A1 WO 2018112514 A1 WO2018112514 A1 WO 2018112514A1 AU 2017051388 W AU2017051388 W AU 2017051388W WO 2018112514 A1 WO2018112514 A1 WO 2018112514A1
Authority
WO
WIPO (PCT)
Prior art keywords
model
training images
machine learning
learning model
specialised
Prior art date
Application number
PCT/AU2017/051388
Other languages
English (en)
Inventor
Chris MCCOOL
Original Assignee
Queensland University Of Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Priority claimed from AU2016905352A external-priority patent/AU2016905352A0/en
Application filed by Queensland University Of Technology filed Critical Queensland University Of Technology
Publication of WO2018112514A1 publication Critical patent/WO2018112514A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T1/00—General purpose image data processing
    • G06T1/20—Processor architectures; Processor configuration, e.g. pipelining
    • 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

Definitions

  • the non-specialised (generic) deep model has been trained with nonspecific (generic) image data.
  • the non-specific (generic) image data does not include data relating to classifiers of the training images.
  • the compressed model may be used in a computer vision task of a robot.
  • the compressed model may, for example, be configured for weed classification or crop segmentation.
  • the invention resides broadly in a system for generating a compressed machine learning model including:
  • a memory coupled to the processor, the memory including instruction code executable by the processor for:
  • the system further includes a camera, for capturing an image, wherein the instruction code is further executable to classify the captured image according to the compressed machine learning model.
  • the system includes one or more actuators, configured to act based upon classification of the captured image.
  • the actuators may form part of or be associated with a robotic arm, a spray nozzle, a harvesting tool or a mechanical implement.
  • the present invention resides in a system for classifying images using a model generated according to the first or second aspects.
  • Figure 1 illustrates a schematic of a deep learning model generation system, for use in computer vision, according to an embodiment of the present invention
  • Figure 2 illustrates a schematic of an image classification system, according to an embodiment of the present invention
  • Figure 5 illustrates a method of generating a specialised deep model by adapting a generic model, according to an embodiment of the present invention
  • Figure 7 illustrates a computing device, according to an embodiment of the present invention.
  • the generic deep model 105 is trained using images having a wide variety of classifications, such as images of various types of animals, devices, foods, and objects.
  • the generic deep model 105 has not, however, been trained to the specific task in question, such as weed detection.
  • the generic deep model 105 is adapted (fined tuned) using high resolution images 110 to become a specialised deep model 115, and then compressed using low resolution images 120, as outlined below, to form an output model 125.
  • the output model 125 may then be used in classification and segmentation tasks.
  • the generic deep model 105 is adapted (fine-tuned) by replacing a final decision layer of the model with a randomly initialised layer that includes the number of output classes C desired by the specialised model 115.
  • the number of output classes C also corresponds to the number of classes in the high resolution training images 110. This enables the high resolution training images 110 to train the model.
  • a model that is adapted to detect several different weed types may have a final decision layer of the model replaced by a decision layer including an output class corresponding to each weed type.
  • the high resolution training images 110 also include training data for each of the output classes, and thus weed types.
  • the model with the replaced final decision layer is then retrained using the high resolution images 110, to update at least the new final decision layer.
  • parts of the model may be fixed during retraining, to prevent the image data from influencing certain aspects of the model too much. In such case, it is generally desirable to fix layers furthest away from the final decision layer.
  • weights may be applied to parameters defining how much training applies to each of the parameters. For example, layers closer to the final decision layer may be weighted such that they are most influenced during training, whereas layers further from the final decision layer are least influenced.
  • the replaced layers need not be randomly initialised.
  • the replaced layers may be initialised using default settings, or settings specific to the classification task of the model.
  • the specialised deep model 115 is compressed to generate the output model 120.
  • This process is also referred to as distillation and results in an output model 125 that is simpler than the generic deep model 105, and thus better suited for low complexity tasks, such as classification on portable computing devices.
  • the output model 120 will include fewer layers and/or parameters than the specialised deep model 115.
  • the output model 125 is generated using low resolution images that correspond to the high resolution images, the logit outputs of the specialised deep model 115, z, as well as the true class labels, y, that correspond to both the high and low resolution images.
  • the ability of the output model to replicate the true class labels, y, and logit outputs (of the specialised deep model), z, is determined by a loss function.
  • the loss function for producing the correct class labels and the loss function for replicating the logit outputs can then be combined to train the system. This ensures that the system learns how to replicate the output of the complex specialised deep model while still achieving high classification accuracy.
  • the specialised deep model 115 and the output model 125 include outputs at each of the layers, and an activation layer, to assist in interpreting the outputs as probabilities.
  • the outputs may include logit outputs, which are logarithms of predicted probabilities of each of the classes, and the activation layer may comprise a softmax activation function to convert the logit outputs to probabilities.
  • the output model 125 By training the output model 125 using the logit outputs generated by specialised deep model 115, rather than the output probabilities of the activation layer, the output model 125 is able to not only learn from the ultimate output of the specialised deep model 115, but also the internal workings of the model. In short, such training does not suffer from the information loss that occurs from passing through logits to probability.
  • the low resolution training images 120 may be down-sampled from the high resolution training images 110.
  • the high resolution training images 110 may be up-sampled from the low resolution training images 120.
  • the high and low resolution training images may be generated from another common image.
  • the high resolution training images 110 may be 120x120 pixels in size
  • the low resolution training images 120 may be 81x81 pixels in size.
  • Figure 2 illustrates a schematic of an image classification system 200, according to an embodiment of the present invention.
  • the prediction 215 comprises an output of the model 205, and may comprise data relating to probabilities of each of the possible classes (e.g. 0.1 and 0.9 for two classes), or a class label (e.g. "weed").
  • the input image 210 is generated by applying a sliding window to a larger image.
  • a sliding window corresponding to the size of the low resolution training images 120 may be used to generate a plurality of input images 210 corresponding to a plurality of regions of the larger image. This enables the system 200 to classify regions of an image.
  • FIG. 3 illustrates a schematic of an image classification system 300, according to a further embodiment of the present invention.
  • the image classification system 300 is similar to the image classification system 200, and includes an input image 210, for classification, but a plurality of models 305.
  • Each model 305 may be generated using the system 100, but using training images in a different order, or otherwise adding variation into the system. As such, each model 305 has been generated to perform the same task as the other models 305.
  • a generic (non-specific) model is received.
  • the generic model may comprise the GoogLeNet model, which has been trained using a large number of images illustrating various classes.
  • specialised training images are received.
  • the specialised training images relate to a specialised task, such as weed detection, and may include one or more different classes (e.g. different types of weeds).
  • a specialised deep model is generated by adapting the generic model using the training images.
  • the specialised deep model is able to classify images relating to the specialised task (e.g. weed classification). Further detail of how the specialised deep model is generated is provided below.
  • low resolution versions of the specialised training images are generated. Any suitable sampling algorithm may be used, including bilinear and bicubic resampling algorithms, to generate the low resolution versions of the specialised training images.
  • a compressed model is generated using the specialised deep model and the low resolution training images.
  • the compressed model is also able to classify images relating to the specialised task (e.g. weed classification), and is trained to mimic the decisions of the specialised deep model. Further detail of how the compressed model is generated is provided below.
  • a new image is received for classification. Unlike the training images, the new image is not associated with any class label y.
  • the new image may, for example, be received from a camera of a classification device, or be part of a larger image captured.
  • the new image is classified using the compressed model.
  • the new image is classified using probabilities output from the compressed model, wherein a classification having a highest probability is chosen.
  • Steps 405-425 relate to the generation of the compressed model
  • steps 430-435 relate to use of the compressed model.
  • the compressed model may be generated once only, and using different hardware than the compressed model. For example, steps 405-425 may be performed off-site, whereas the steps 430-435 may be performed on-site.
  • the specialised deep model may be trained from scratch, i.e. without adapting the generic model. This is useful if there are sufficient training images to train the specialised deep model and/or if no suitable generic model is available. In such case, steps 405-415 may be replaced by a step including receiving the specialised deep model.
  • Figure 5 illustrates a method 500 of generating a specialised deep model by adapting a generic model, according to an embodiment of the present invention.
  • the method 500 may be similar or identical to that used in step 415 of the method 400, for example.
  • a new output layer of a deep model is generated, where output of the new layer correspond to classes in the training images.
  • the outputs may correspond to each weed which may be classified.
  • a final decision layer of the generic model is replaced with the new output layer generated at step 505.
  • outputs of the earlier layers of the model are coupled to the new final decision layer.
  • the model including the new output layer, is trained using training images.
  • the new layer is updated, and potentially other layers of the model.
  • the layers closest to the new layer are updated more than the layers further from the model, which ensures that the model is trained to recognise the specialised data, while retaining general image classification features from the generic model.
  • the training process may include incrementally updating the layers.
  • the training images may be used to update the model, after which the training images are provided to the updated model, and so on, until a desired level of training is reached.
  • Figure 6 illustrates a method 600 of compressing a specialised deep model, according to an embodiment of the present invention.
  • the method 500 may be similar or identical to that used in step 425 of the method 400, for example.
  • a base model is generated as a starting point for the compressed model.
  • the base model includes the same number of outputs as the specialised deep model in which it will mimic, but fewer layers and parameters.
  • step 610 low resolution images are provided to the base model, the low resolution image corresponding to a high resolution images on which the specialised deep model was trained.
  • logit outputs of the base model generated using the low resolution images are compared with logit outputs of the deep model with the corresponding high resolution image.
  • the base model is updated based upon the difference in logit outputs. This enables the base model to mimic the specialised deep model, as described above.
  • Steps 610-620 are then repeated until a desired level of training is reached, upon which the base model mimics the specialised deep model.
  • the database 720 generally includes a plurality of training images, and
  • the training images enable the models to be adapted to suit specific data of the training images.
  • the processor 705 is further coupled to a data interface 725, which may in turn be coupled to a monitor and/or data input device, to enable a user to control the computing device 700.
  • the data interface may also be coupled to input-output devices, such as portable memories, disk drives and the like.
  • a network interface 730 is coupled to the processor to enable network based input, output and control.
  • the computing device may be configured to retrieve test images from an external data source on the Internet.
  • the robotic arm may incorporate (or be replaced by) a spray nozzle (in case of herbicide based weed eradication), a harvesting tool (in case of a harvesting robot), mechanical implements (in case of mechanical weed destruction), or any other suitable tool or implement.
  • a spray nozzle in case of herbicide based weed eradication
  • a harvesting tool in case of a harvesting robot
  • mechanical implements in case of mechanical weed destruction
  • computing device 700 may readily be adapted to suit other purposes, such as medical imaging, where blood cells, tissue and/or other material is classified, pedestrian recognition in the context of vehicle safety, classification of fruit or vegetables, or any other suitable classification or identification task.
  • a weed classification model generated according to the method 400 of Figure 4 was assessed for three weeds: volunteer cotton, sow thistle and wild oats. These weeds are herbicide resistant and of importance for Queensland.
  • the training images were taken in a field, and validation images were taken in a similar field, at a similar time.
  • the evaluation image set is completely separate data that was captured four months later in a similar field.
  • the classification was based on detecting a region of interest (Rol) and then determining a class for the Rol.
  • Adapted-IV3 corresponds to a specialised deep model according to the method 500 of Figure 5 obtained by adapting a generic model.
  • WeedNet- l corresponds to the method 500 of Figure 5, where the specialised deep model is trained from scratch (i.e. without adapting a generic model).
  • AgNet corresponds to method 400 of Figure 4 where the specialised deep model is Adapted-IV3.
  • LBP RF relates to local binary patterns (Ojala et al., 2002) with random forest.
  • the LBP RF system achieves an accuracy of 87.7% for the validation set and 83.9% for the test set. This is lower than the worst performing deep learnt feature, i.e. where the specialised deep model is trained from scratch, which achieves an accuracy of 97.1% for the validation set and 86.4% for the test set.
  • Training a compressed model (AgNet) corresponding to the method 400 of Figure 4, leads to a more accurate model than training from scratch as shown by the results for AgNet vs WeedNet-vl where AgNet achieves an accuracy of 97.9% for the validation set and 89.8% for the evaluation set.
  • Adapted-IV3 and WeedNet-vl perform the best, however training from scratch (WeedNet-vl) doesn't perform as well as fine-tuning a well trained model (Adapted-IV3).
  • Training a compressed model (AgNet) leads to a more accurate model than training from scratch as shown by the results for AgNet and WeedNet-v 1.
  • AgNet is about an order of magnitude faster than Adapted-IV3, thus providing a trade off between speed and complexity against accuracy.
  • a weed segmentation model was generated according to the method 400 of Figure 4 and was assessed on the Crop/Weed Field Image Dataset (CWFID) (Haug and Ostermann, 2014).
  • CWFID Crop/Weed Field Image Dataset
  • each pixel was classified as either crop or weed by extracting a region of interest (Rol) around the pixel.
  • Rol region of interest
  • Adapted-IV3 corresponds to the specialised deep model according to the method 500 of Figure 5.
  • deep networks i.e. Adapted-IV3, AgNet, Mix-AgNet,
  • Minilnception and Mix-Minilnception outperform handcrafted methods (LBP and Shape+Stat.).
  • Minilnception and AgNet correspond to method 400 of Figure 4 using different network architectures (Minilnception and AgNet).
  • Mix-Minilnception and Mix-AgNet correspond to a combination of N compressed models corresponding to the system 300 of Figure 3.
  • Using a mixture (Mix-AgNet or Mix-Minilnception) or combination of compressed models (AgNet or Minilnception) improves the accuracy of the system but also increases the complexity (number of parameters) and decreases the speed of the system, as can be seen in Table II.
  • a weed segmentation model was generated according to the method 400 of Figure 4 and was assessed in relation to capsicum (sweet pepper) segmentation on the Sweet Pepper Dataset (McCool et al., 2016). In this case, each pixel was classified as being capsicum or not by extracting a region of interest (Rol) around the pixel.
  • Adapted-IV3, Minilnception and Mix- Minilnception outperform the two baseline methods (McCool et al., 2016), which comprise a combination of three visual features and colour (Baseline-Fusion), and the single best visual feature (Baseline-LBP).
  • Minilnception and AgNet correspond to the method 400 of Figure 4
  • Adapted-IV3 corresponds to the specialised deep model according to the method 500 of Figure 5.
  • Mix-Minilnception corresponds to a combination of N compressed models (Minilnception) corresponding to the system 300 of Figure 3.
  • embodiments of the present invention provide improvements in computational efficiency (speed) by more than an order of magnitude.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Biophysics (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Image Analysis (AREA)

Abstract

L'invention concerne un procédé et un système de génération d'un modèle d'apprentissage automatique compressé. Le procédé comprend les étapes consistant : à fournir un premier ensemble d'images d'apprentissage à un premier modèle d'apprentissage automatique pour générer un premier ensemble de sorties ; à fournir un second ensemble d'images d'apprentissage à un second modèle d'apprentissage automatique pour générer un second ensemble de sorties, le second ensemble d'images d'apprentissage correspondant au premier ensemble d'images d'apprentissage ayant une résolution inférieure ; et à mettre à jour le second modèle d'apprentissage automatique en fonction d'une différence entre les premier et second ensembles de sorties pour générer le modèle d'apprentissage automatique compressé.
PCT/AU2017/051388 2016-12-23 2017-12-14 Systèmes et procédés d'apprentissage profond destinés à être utilisés dans la vision artificielle WO2018112514A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
AU2016905352A AU2016905352A0 (en) 2016-12-23 Deep learning systems and methods for use in computer vision
AU2016905352 2016-12-23

Publications (1)

Publication Number Publication Date
WO2018112514A1 true WO2018112514A1 (fr) 2018-06-28

Family

ID=62624306

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/AU2017/051388 WO2018112514A1 (fr) 2016-12-23 2017-12-14 Systèmes et procédés d'apprentissage profond destinés à être utilisés dans la vision artificielle

Country Status (1)

Country Link
WO (1) WO2018112514A1 (fr)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110334642A (zh) * 2019-07-01 2019-10-15 河南牧业经济学院 一种猪行为的机器视觉识别方法及系统
WO2020032661A1 (fr) * 2018-08-10 2020-02-13 Samsung Electronics Co., Ltd. Appareil électronique, son procédé de commande, et procédé de commande d'un serveur
EP3809366A1 (fr) 2019-10-15 2021-04-21 Aisapack Holding SA Procédé de fabrication
US20220207275A1 (en) * 2020-12-30 2022-06-30 Zoox, Inc. Multi-resolution top-down prediction
WO2023086585A1 (fr) * 2021-11-12 2023-05-19 Covera Health Auto-influence repondérée pour élimination du bruit d'étiquetage dans des données d'imagerie médicale

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4175860A (en) * 1977-05-31 1979-11-27 Rush-Presbyterian-St. Luke's Medical Center Dual resolution method and apparatus for use in automated classification of pap smear and other samples
US5331550A (en) * 1991-03-05 1994-07-19 E. I. Du Pont De Nemours And Company Application of neural networks as an aid in medical diagnosis and general anomaly detection
WO2000036524A1 (fr) * 1998-12-16 2000-06-22 Sarnoff Corporation Procede et appareil d'entrainement d'un reseau neuronal a la detection d'objets dans une image
US20030233335A1 (en) * 2002-06-17 2003-12-18 Mims Aj Student neural network
US9129190B1 (en) * 2013-12-04 2015-09-08 Google Inc. Identifying objects in images
US20170200063A1 (en) * 2016-01-13 2017-07-13 Ford Global Technologies, Llc Low- and high-fidelity classifiers applied to road-scene images

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4175860A (en) * 1977-05-31 1979-11-27 Rush-Presbyterian-St. Luke's Medical Center Dual resolution method and apparatus for use in automated classification of pap smear and other samples
US5331550A (en) * 1991-03-05 1994-07-19 E. I. Du Pont De Nemours And Company Application of neural networks as an aid in medical diagnosis and general anomaly detection
WO2000036524A1 (fr) * 1998-12-16 2000-06-22 Sarnoff Corporation Procede et appareil d'entrainement d'un reseau neuronal a la detection d'objets dans une image
US20030233335A1 (en) * 2002-06-17 2003-12-18 Mims Aj Student neural network
US9129190B1 (en) * 2013-12-04 2015-09-08 Google Inc. Identifying objects in images
US20170200063A1 (en) * 2016-01-13 2017-07-13 Ford Global Technologies, Llc Low- and high-fidelity classifiers applied to road-scene images

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
MCCOOL, C. ET AL.: "Mixtures of Lightweight Deep Convolutional Neural Networks: Applied to Agricultural Robotics", IEEE ROBOTICS AND AUTOMATION LETTERS, vol. 2, no. 3, July 2017 (2017-07-01), pages 1344 - 1351, XP055496006, [retrieved on 20170209] *
ZANGENEH. E. ET AL.: "Low Resolution Face Recognition Using a Two-Branch Deep Convolutional Neural Network Architecture", ONLINE JOURNA, 20 June 2017 (2017-06-20), XP080771113, Retrieved from the Internet <URL:https://arxiv.org/abs/1706.06247> [retrieved on 20180201] *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020032661A1 (fr) * 2018-08-10 2020-02-13 Samsung Electronics Co., Ltd. Appareil électronique, son procédé de commande, et procédé de commande d'un serveur
US11388465B2 (en) 2018-08-10 2022-07-12 Samsung Electronics Co., Ltd. Electronic apparatus and method for upscaling a down-scaled image by selecting an improved filter set for an artificial intelligence model
US11825033B2 (en) 2018-08-10 2023-11-21 Samsung Electronics Co., Ltd. Apparatus and method with artificial intelligence for scaling image data
CN110334642A (zh) * 2019-07-01 2019-10-15 河南牧业经济学院 一种猪行为的机器视觉识别方法及系统
EP3809366A1 (fr) 2019-10-15 2021-04-21 Aisapack Holding SA Procédé de fabrication
WO2021074708A1 (fr) 2019-10-15 2021-04-22 Aisapack Holding Sa Procédé de fabrication
US20220207275A1 (en) * 2020-12-30 2022-06-30 Zoox, Inc. Multi-resolution top-down prediction
US11847831B2 (en) * 2020-12-30 2023-12-19 Zoox, Inc. Multi-resolution top-down prediction
WO2023086585A1 (fr) * 2021-11-12 2023-05-19 Covera Health Auto-influence repondérée pour élimination du bruit d'étiquetage dans des données d'imagerie médicale

Similar Documents

Publication Publication Date Title
Bosilj et al. Transfer learning between crop types for semantic segmentation of crops versus weeds in precision agriculture
WO2018112514A1 (fr) Systèmes et procédés d&#39;apprentissage profond destinés à être utilisés dans la vision artificielle
Häni et al. A comparative study of fruit detection and counting methods for yield mapping in apple orchards
Azizah et al. Deep learning implementation using convolutional neural network in mangosteen surface defect detection
Baccouche et al. Sequential deep learning for human action recognition
US20190164047A1 (en) Object recognition using a convolutional neural network trained by principal component analysis and repeated spectral clustering
Wang et al. YOLOv3‐Litchi Detection Method of Densely Distributed Litchi in Large Vision Scenes
Škrabánek et al. Detection of grapes in natural environment using support vector machine classifier
Banerjee et al. Automated Diagnosis of Marigold Leaf Diseases using a Hybrid CNN-SVM Model
Omer et al. Plant disease diagnosing based on deep learning techniques
Suwais et al. A review on classification methods for plants leaves recognition
Pasquale et al. Real-world object recognition with off-the-shelf deep conv nets: How many objects can icub learn?
Alashhab et al. Precise ship location with cnn filter selection from optical aerial images
Almalky et al. An Efficient Deep Learning Technique for Detecting and Classifying the Growth of Weeds on Fields
Dolezel et al. Detection of grapes in natural environment using feedforward neural network as a classifier
Popescu et al. Comparative Study of Neural Networks Used in Halyomorpha Halys Detection
Solberg Deep neural networks for object detection in agricultural robotics
Chaudhary et al. An efficient approach for automated system to identify the rice crop disease using intensity level based multi-fractal dimension and twin support vector machine
Chen et al. Inductive Clustering: automating low-level segmentation in high resolution images
Sahu et al. CNN based disease detection in Apple Leaf via transfer learning
Shobana et al. Fruit Freshness Detecting System Using Deep Learning and Raspberry PI
Kulkarni et al. CoffeeGAN: An Effective Data Augmentation Model for Coffee Plant Diseases
Baruque et al. WeVoS scale invariant map
Hi et al. Lantana Camara Flower Detection Using an Improved Lightweight Convolutional Neural Networks in YOLOv5
Balakrishnan et al. Computing WHERE-WHAT classification through FLIKM and deep learning algorithms

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17884426

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17884426

Country of ref document: EP

Kind code of ref document: A1