Storing Encoded Episodes as Concepts for Continual Learning

By Ali Ayub, Alan R. Wagner. 2020

13 citations recorded by the source corpus.

The two main challenges faced by continual learning approaches are catastrophic forgetting and memory limitations on the storage of data. To cope with these challenges, we propose a novel, cognitively-inspired approach which trains autoencoders with Neural Style Transfer to encode and store images. Reconstructed images from encoded episodes are replayed when training the classifier model on a new task to avoid catastrophic forgetting. The loss function for the reconstructed images is weighted to reduce its effect during classifier training to cope with image degradation. When the system runs out of memory the encoded episodes are converted into centroids and covariance matrices, which are used to generate pseudo-images during classifier training, keeping classifier performance stable with less memory. Our approach increases classification accuracy by 13-17% over state-of-the-art methods on benchmark datasets, while requiring 78% less storage space.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Online Learning of Objects through Curiosity-Driven Active Learning (2021, background): “To overcome catastrophic forgetting, many different deep learning techniques have been proposed for continual learning [12], [13], [14], [15], [16], [17], [18].”
Tell me what this is: Few-Shot Incremental Object Learning by a Robot (2020, background): “To avoid storage of real samples, some approaches use generative-memory and regenerate samples of old classes using GANs or autoencoders [9], [10], [11], however the performance of these approaches is generally inferior to approaches that store real images.”
F-SIOL-310: A Robotic Dataset and Benchmark for Few-Shot Incremental Object Learning (2021, methodology): “Further, most current incremental learning approaches [5], [6], [7], [8] are not designed for few-shot incremental learning (FSIL),”
Centroid Based Concept Learning for RGB-D Indoor Scene Classification (2019, background): “Furthermore, it can applied to other classiﬁcation tasks, such as incremental learning of objects from object-centric RGB images [2, 3, 4].”