Memory Augmented Neural Network with Gaussian Embeddings for One-Shot Learning

Memory augmented neural networks that use pointwise embeddings have been successfully applied to one-shot learning, however, Gaussian embeddings are more versatile and provide an opportunity to build models capturing latent structure. As a step towards combining both, we construct a memory augmented network with Gaussian embeddings. We provide results of one-shot classification on Omniglot and LFW-a datasets, and since the resulting model is generative, image reconstruction results on Omniglot. Additionally, we explain how to learn more classes in one-shot using memory augmented neural networks with a method that does not depend on the type of embeddings.

[1]  James L. McClelland,et al.  What Learning Systems do Intelligent Agents Need? Complementary Learning Systems Theory Updated , 2016, Trends in Cognitive Sciences.

[2]  Joshua B. Tenenbaum,et al.  Human-level concept learning through probabilistic program induction , 2015, Science.

[3]  Ambedkar Dukkipati,et al.  Generative Adversarial Residual Pairwise Networks for One Shot Learning , 2017, ArXiv.

[4]  Ole Winther,et al.  Auxiliary Deep Generative Models , 2016, ICML.

[5]  Richard S. Zemel,et al.  Prototypical Networks for Few-shot Learning , 2017, NIPS.

[6]  Jörg Bornschein,et al.  Variational Memory Addressing in Generative Models , 2017, NIPS.

[7]  Ryan P. Adams,et al.  Composing graphical models with neural networks for structured representations and fast inference , 2016, NIPS.

[8]  Jiwon Kim,et al.  Continual Learning with Deep Generative Replay , 2017, NIPS.

[9]  Tal Hassner,et al.  Effective Unconstrained Face Recognition by Combining Multiple Descriptors and Learned Background Statistics , 2011, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[10]  E. Hoogeboom.  Few-shot Classification by Learning Disentangled Representations , 2017 .

[11]  Bartunov Sergey,et al.  Meta-Learning with Memory-Augmented Neural Networks , 2016 .

[12]  Gregory R. Koch,et al.  Siamese Neural Networks for One-Shot Image Recognition , 2015 .

[13]  Yann LeCun,et al.  Disentangling factors of variation in deep representation using adversarial training , 2016, NIPS.

[14]  Navdeep Jaitly,et al.  Adversarial Autoencoders , 2015, ArXiv.

[15]  Eric Eaton,et al.  Online Contrastive Divergence with Generative Replay: Experience Replay without Storing Data , 2016, ArXiv.

[16]  Aurko Roy,et al.  Learning to Remember Rare Events , 2017, ICLR.

[17]  James Philbin,et al.  FaceNet: A unified embedding for face recognition and clustering , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[18]  A.R. Runnalls,et al.  A Kullback-Leibler Approach to Gaussian Mixture Reduction , 2007 .

[19]  Alex Graves,et al.  Neural Turing Machines , 2014, ArXiv.