S-prompts learning with pre-trained transformers: An Occam's razor for domain incremental learning

State-of-the-art deep neural networks are still struggling to address the catastrophic forgetting problem in continual learning. In this paper, we propose one simple paradigm (named as S-Prompting) and two concrete approaches to highly reduce the forgetting degree in one of the most typical continua...

Full description

Saved in:
Bibliographic Details
Main Authors: WANG, Yabin, HUANG, Zhiwu, HONG, Xiaopeng.
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2022
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/7614
https://ink.library.smu.edu.sg/context/sis_research/article/8617/viewcontent/03_Sprompts_NeurIPS2022.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
id sg-smu-ink.sis_research-8617
record_format dspace
spelling sg-smu-ink.sis_research-86172022-12-22T03:26:02Z S-prompts learning with pre-trained transformers: An Occam's razor for domain incremental learning WANG, Yabin HUANG, Zhiwu HONG, Xiaopeng. State-of-the-art deep neural networks are still struggling to address the catastrophic forgetting problem in continual learning. In this paper, we propose one simple paradigm (named as S-Prompting) and two concrete approaches to highly reduce the forgetting degree in one of the most typical continual learning scenarios, i.e., domain increment learning (DIL). The key idea of the paradigm is to learn prompts independently across domains with pre-trained transformers, avoiding the use of exemplars that commonly appear in conventional methods. This results in a win-win game where the prompting can achieve the best for each domain. The independent prompting across domains only requests one single cross-entropy loss for training and one simple K-NN operation as a domain identifier for inference. The learning paradigm derives an image prompt learning approach and a novel language-image prompt learning approach. Owning an excellent scalability (0.03% parameter increase per domain), the best of our approaches achieves a remarkable relative improvement (an average of about 30%) over the best of the state-of-the-art exemplar-free methods for three standard DIL tasks, and even surpasses the best of them relatively by about 6% in average when they use exemplars. Source code is available at https://github.com/iamwangyabin/S-Prompts. 2022-12-01T08:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/7614 https://ink.library.smu.edu.sg/context/sis_research/article/8617/viewcontent/03_Sprompts_NeurIPS2022.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Prompts Learning Pre-trained Transformers Occam's Razor Domain Incremental Learning Artificial Intelligence and Robotics Databases and Information Systems
institution Singapore Management University
building SMU Libraries
continent Asia
country Singapore
Singapore
content_provider SMU Libraries
collection InK@SMU
language English
topic Prompts Learning
Pre-trained Transformers
Occam's Razor
Domain Incremental Learning
Artificial Intelligence and Robotics
Databases and Information Systems
spellingShingle Prompts Learning
Pre-trained Transformers
Occam's Razor
Domain Incremental Learning
Artificial Intelligence and Robotics
Databases and Information Systems
WANG, Yabin
HUANG, Zhiwu
HONG, Xiaopeng.
S-prompts learning with pre-trained transformers: An Occam's razor for domain incremental learning
description State-of-the-art deep neural networks are still struggling to address the catastrophic forgetting problem in continual learning. In this paper, we propose one simple paradigm (named as S-Prompting) and two concrete approaches to highly reduce the forgetting degree in one of the most typical continual learning scenarios, i.e., domain increment learning (DIL). The key idea of the paradigm is to learn prompts independently across domains with pre-trained transformers, avoiding the use of exemplars that commonly appear in conventional methods. This results in a win-win game where the prompting can achieve the best for each domain. The independent prompting across domains only requests one single cross-entropy loss for training and one simple K-NN operation as a domain identifier for inference. The learning paradigm derives an image prompt learning approach and a novel language-image prompt learning approach. Owning an excellent scalability (0.03% parameter increase per domain), the best of our approaches achieves a remarkable relative improvement (an average of about 30%) over the best of the state-of-the-art exemplar-free methods for three standard DIL tasks, and even surpasses the best of them relatively by about 6% in average when they use exemplars. Source code is available at https://github.com/iamwangyabin/S-Prompts.
format text
author WANG, Yabin
HUANG, Zhiwu
HONG, Xiaopeng.
author_facet WANG, Yabin
HUANG, Zhiwu
HONG, Xiaopeng.
author_sort WANG, Yabin
title S-prompts learning with pre-trained transformers: An Occam's razor for domain incremental learning
title_short S-prompts learning with pre-trained transformers: An Occam's razor for domain incremental learning
title_full S-prompts learning with pre-trained transformers: An Occam's razor for domain incremental learning
title_fullStr S-prompts learning with pre-trained transformers: An Occam's razor for domain incremental learning
title_full_unstemmed S-prompts learning with pre-trained transformers: An Occam's razor for domain incremental learning
title_sort s-prompts learning with pre-trained transformers: an occam's razor for domain incremental learning
publisher Institutional Knowledge at Singapore Management University
publishDate 2022
url https://ink.library.smu.edu.sg/sis_research/7614
https://ink.library.smu.edu.sg/context/sis_research/article/8617/viewcontent/03_Sprompts_NeurIPS2022.pdf
_version_ 1770576394667950080