Learning to Remember Rare Events

9 Mar 2017  ·  Łukasz Kaiser, Ofir Nachum, Aurko Roy, Samy Bengio ·

Despite recent advances, memory-augmented deep neural networks are still limited when it comes to life-long and one-shot learning, especially in remembering rare events. We present a large-scale life-long memory module for use in deep learning... The module exploits fast nearest-neighbor algorithms for efficiency and thus scales to large memory sizes. Except for the nearest-neighbor query, the module is fully differentiable and trained end-to-end with no extra supervision. It operates in a life-long manner, i.e., without the need to reset it during training. Our memory module can be easily added to any part of a supervised neural network. To show its versatility we add it to a number of networks, from simple convolutional ones tested on image classification to deep sequence-to-sequence and recurrent-convolutional models. In all cases, the enhanced network gains the ability to remember and do life-long one-shot learning. Our module remembers training examples shown many thousands of steps in the past and it can successfully generalize from them. We set new state-of-the-art for one-shot learning on the Omniglot dataset and demonstrate, for the first time, life-long one-shot learning in recurrent neural networks on a large-scale machine translation task. read more

PDF Abstract

Datasets


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Few-Shot Image Classification OMNIGLOT - 1-Shot, 20-way ConvNet with Memory Module Accuracy 95% # 13
Few-Shot Image Classification OMNIGLOT - 1-Shot, 5-way ConvNet with Memory Module Accuracy 98.4 # 13
Few-Shot Image Classification OMNIGLOT - 5-Shot, 20-way ConvNet with Memory Module Accuracy 98.6% # 10
Few-Shot Image Classification OMNIGLOT - 5-Shot, 5-way ConvNet with Memory Module Accuracy 99.6 # 10

Methods


No methods listed for this paper. Add relevant methods here