Gradient Correlation Subspace Learning against Catastrophic Forgetting

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dubnov, Tammuz, Thengane, Vishal
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914701300989952
author Dubnov, Tammuz
Thengane, Vishal
author_facet Dubnov, Tammuz
Thengane, Vishal
contents Efficient continual learning techniques have been a topic of significant research over the last few years. A fundamental problem with such learning is severe degradation of performance on previously learned tasks, known also as catastrophic forgetting. This paper introduces a novel method to reduce catastrophic forgetting in the context of incremental class learning called Gradient Correlation Subspace Learning (GCSL). The method detects a subspace of the weights that is least affected by previous tasks and projects the weights to train for the new task into said subspace. The method can be applied to one or more layers of a given network architectures and the size of the subspace used can be altered from layer to layer and task to task. Code will be available at \href{https://github.com/vgthengane/GCSL}{https://github.com/vgthengane/GCSL}
format Preprint
id arxiv_https___arxiv_org_abs_2403_02334
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Gradient Correlation Subspace Learning against Catastrophic Forgetting
Dubnov, Tammuz
Thengane, Vishal
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
Efficient continual learning techniques have been a topic of significant research over the last few years. A fundamental problem with such learning is severe degradation of performance on previously learned tasks, known also as catastrophic forgetting. This paper introduces a novel method to reduce catastrophic forgetting in the context of incremental class learning called Gradient Correlation Subspace Learning (GCSL). The method detects a subspace of the weights that is least affected by previous tasks and projects the weights to train for the new task into said subspace. The method can be applied to one or more layers of a given network architectures and the size of the subspace used can be altered from layer to layer and task to task. Code will be available at \href{https://github.com/vgthengane/GCSL}{https://github.com/vgthengane/GCSL}
title Gradient Correlation Subspace Learning against Catastrophic Forgetting
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2403.02334