Move-to-Data: A new Continual Learning approach with Deep CNNs, Application for image-class recognition

In many real-life tasks of application of supervised learning approaches, all the training data are not available at the same time. The examples are lifelong image classification or recognition of environmental objects during interaction of instrumented persons with their environment, enrichment of an online-database with more images. It is necessary to pre-train the model at a "training recording phase" and then adjust it to the new coming data. This is the task of incremental/continual learning approaches. Amongst different problems to be solved by these approaches such as introduction of new categories in the model, refining existing categories to sub-categories and extending trained classifiers over them, ... we focus on the problem of adjusting pre-trained model with new additional training data for existing categories. We propose a fast continual learning layer at the end of the neuronal network. Obtained results are illustrated on the opensource CIFAR benchmark dataset. The proposed scheme yields similar performances as retraining but with drastically lower computational cost.

[1]  Cordelia Schmid,et al.  Incremental Learning of Object Detectors without Catastrophic Forgetting , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[2]  Jenny Benois-Pineau,et al.  Search of objects of interest in videos , 2012, 2012 10th International Workshop on Content-Based Multimedia Indexing (CBMI).

[3]  Yoshua Bengio,et al.  How transferable are features in deep neural networks? , 2014, NIPS.

[4]  Jenny Benois-Pineau,et al.  Perceptually-guided Understanding of Egocentric Video Content: Recognition of Objects to Grasp , 2018, ICMR.

[5]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[6]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[7]  Alex Krizhevsky,et al.  Learning Multiple Layers of Features from Tiny Images , 2009 .

[8]  Jenny Benois-Pineau,et al.  Introduction of Explicit Visual Saliency in Training of Deep CNNs: Application to Architectural Styles Classification , 2018, 2018 International Conference on Content-Based Multimedia Indexing (CBMI).

[9]  Tim Salimans,et al.  Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks , 2016, NIPS.

[10]  Michèle Sebag,et al.  Distributed and Incremental Clustering Based on Weighted Affinity Propagation , 2008, STAIRS.

[11]  Jenny Benois-Pineau,et al.  Perceptually-guided deep neural networks for ego-action prediction: Object grasping , 2019, Pattern Recognit..

[12]  Geoffrey E. Hinton,et al.  Learning internal representations by error propagation , 1986 .

[13]  Yi Li,et al.  DeepTrack: Learning Discriminative Feature Representations by Convolutional Neural Networks for Visual Tracking , 2014, BMVC.

[14]  Jenny Benois-Pineau,et al.  Classification of Alzheimer Disease on Imaging Modalities with Deep CNNs Using Cross-Modal Transfer Learning , 2018, 2018 IEEE 31st International Symposium on Computer-Based Medical Systems (CBMS).

[15]  Nikos Komodakis,et al.  Dynamic Few-Shot Visual Learning Without Forgetting , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[16]  Ting Liu,et al.  Recent advances in convolutional neural networks , 2015, Pattern Recognit..

[17]  James Lee Hafner,et al.  Efficient Color Histogram Indexing for Quadratic Form Distance Functions , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[18]  Nikhil Ketkar,et al.  Deep Learning with Python , 2017 .

[19]  Jian Sun,et al.  Identity Mappings in Deep Residual Networks , 2016, ECCV.

[20]  Jenny Benois-Pineau,et al.  Deep Learning in Mining of Visual Content , 2020, Springer Briefs in Computer Science.

[21]  Stefan Wermter,et al.  Continual Lifelong Learning with Neural Networks: A Review , 2019, Neural Networks.

[22]  Edwin Lughofer,et al.  Extensions of vector quantization for incremental clustering , 2008, Pattern Recognit..

[23]  Svetlana Lazebnik,et al.  Piggyback: Adding Multiple Tasks to a Single, Fixed Network by Learning to Mask , 2018, ArXiv.

[24]  Cordelia Schmid,et al.  End-to-End Incremental Learning , 2018, ECCV.

[25]  Denis Pellerin,et al.  Improving Hierarchical Image Classification with Merged CNN Architectures , 2017, CBMI.