Adaptive aggregation networks for class-incremental learning
Class-Incremental Learning (CIL) aims to learn a classification model with the number of classes increasing phase-by-phase. An inherent problem in CIL is the stability-plasticity dilemma between the learning of old and new classes, i.e., high-plasticity models easily forget old classes, but high-sta...
Saved in:
Main Authors: | , , |
---|---|
Format: | text |
Language: | English |
Published: |
Institutional Knowledge at Singapore Management University
2021
|
Subjects: | |
Online Access: | https://ink.library.smu.edu.sg/sis_research/6119 https://ink.library.smu.edu.sg/context/sis_research/article/7122/viewcontent/CVPR2021_Adaptive_Aggregation_Networks_for_Class_Incremental_Learning__1_.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-7122 |
---|---|
record_format |
dspace |
spelling |
sg-smu-ink.sis_research-71222022-05-18T05:22:45Z Adaptive aggregation networks for class-incremental learning LIU, Yaoyao SCHIELE, Bernt SUN, Qianru Class-Incremental Learning (CIL) aims to learn a classification model with the number of classes increasing phase-by-phase. An inherent problem in CIL is the stability-plasticity dilemma between the learning of old and new classes, i.e., high-plasticity models easily forget old classes, but high-stability models are weak to learn new classes. We alleviate this issue by proposing a novel network architecture called Adaptive Aggregation Networks (AANets) in which we explicitly build two types of residual blocks at each residual level (taking ResNet as the baseline architecture): a stable block and a plastic block. We aggregate the output feature maps from these two blocks and then feed the results to the next-level blocks. We adapt the aggregation weights in order to balance these two types of blocks, i.e., between stability and plasticity, dynamically. We conduct extensive experiments on three CIL benchmarks: CIFAR-100, ImageNet-Subset, and ImageNet, and show that many existing CIL methods can be straightforwardly incorporated into the architecture of AANets to boost their performances. 2021-06-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/6119 info:doi/10.1109/CVPR46437.2021.00257 https://ink.library.smu.edu.sg/context/sis_research/article/7122/viewcontent/CVPR2021_Adaptive_Aggregation_Networks_for_Class_Incremental_Learning__1_.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 Adaptation models Computer vision Adaptive systems Computer architecture Network architecture Benchmark testing Stability analysis Artificial Intelligence and Robotics OS and Networks |
institution |
Singapore Management University |
building |
SMU Libraries |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
SMU Libraries |
collection |
InK@SMU |
language |
English |
topic |
Adaptation models Computer vision Adaptive systems Computer architecture Network architecture Benchmark testing Stability analysis Artificial Intelligence and Robotics OS and Networks |
spellingShingle |
Adaptation models Computer vision Adaptive systems Computer architecture Network architecture Benchmark testing Stability analysis Artificial Intelligence and Robotics OS and Networks LIU, Yaoyao SCHIELE, Bernt SUN, Qianru Adaptive aggregation networks for class-incremental learning |
description |
Class-Incremental Learning (CIL) aims to learn a classification model with the number of classes increasing phase-by-phase. An inherent problem in CIL is the stability-plasticity dilemma between the learning of old and new classes, i.e., high-plasticity models easily forget old classes, but high-stability models are weak to learn new classes. We alleviate this issue by proposing a novel network architecture called Adaptive Aggregation Networks (AANets) in which we explicitly build two types of residual blocks at each residual level (taking ResNet as the baseline architecture): a stable block and a plastic block. We aggregate the output feature maps from these two blocks and then feed the results to the next-level blocks. We adapt the aggregation weights in order to balance these two types of blocks, i.e., between stability and plasticity, dynamically. We conduct extensive experiments on three CIL benchmarks: CIFAR-100, ImageNet-Subset, and ImageNet, and show that many existing CIL methods can be straightforwardly incorporated into the architecture of AANets to boost their performances. |
format |
text |
author |
LIU, Yaoyao SCHIELE, Bernt SUN, Qianru |
author_facet |
LIU, Yaoyao SCHIELE, Bernt SUN, Qianru |
author_sort |
LIU, Yaoyao |
title |
Adaptive aggregation networks for class-incremental learning |
title_short |
Adaptive aggregation networks for class-incremental learning |
title_full |
Adaptive aggregation networks for class-incremental learning |
title_fullStr |
Adaptive aggregation networks for class-incremental learning |
title_full_unstemmed |
Adaptive aggregation networks for class-incremental learning |
title_sort |
adaptive aggregation networks for class-incremental learning |
publisher |
Institutional Knowledge at Singapore Management University |
publishDate |
2021 |
url |
https://ink.library.smu.edu.sg/sis_research/6119 https://ink.library.smu.edu.sg/context/sis_research/article/7122/viewcontent/CVPR2021_Adaptive_Aggregation_Networks_for_Class_Incremental_Learning__1_.pdf |
_version_ |
1770575824796254208 |