Saliency Guided Experience Packing for Replay in Continual Learning

被引:4
|
作者
Saha, Gobinda [1 ]
Roy, Kaushik [1 ]
机构
[1] Purdue Univ, Elmore Family Sch Elect & Comp Engn, W Lafayette, IN 47907 USA
来源
2023 IEEE/CVF WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV) | 2023年
基金
美国国家科学基金会;
关键词
D O I
10.1109/WACV56688.2023.00524
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Artificial learning systems aspire to mimic human intelligence by continually learning from a stream of tasks without forgetting past knowledge. One way to enable such learning is to store past experiences in the form of input examples in episodic memory and replay them when learning new tasks. However, performance of such method suffers as the size of the memory becomes smaller. In this paper, we propose a new approach for experience replay, where we select the past experiences by looking at the saliency maps which provide visual explanations for the model's decision. Guided by these saliency maps, we pack the memory with only the parts or patches of the input images important for the model's prediction. While learning a new task, we replay these memory patches with appropriate zero-padding to remind the model about its past decisions. We evaluate our algorithm on CIFAR-100, miniImageNet and CUB datasets and report better performance than the state-of-the-art approaches. With qualitative and quantitative analyses we show that our method captures richer summaries of past experiences without any memory increase, and hence performs well with small episodic memory.
引用
收藏
页码:5262 / 5272
页数:11
相关论文
共 50 条
  • [41] Generative Feature Replay with Orthogonal Weight Modification for Continual Learning
    Shen, Gehui
    Zhang, Song
    Chen, Xiang
    Deng, Zhi-Hong
    2021 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2021,
  • [42] BinPlay: A Binary Latent Autoencoder for Generative Replay Continual Learning
    Deja, Kamil
    Wawrzynski, Pawel
    Marczak, Daniel
    Masarczyk, Wojciech
    Trzcinski, Tomasz
    2021 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2021,
  • [43] Solving Continual Learning with Noisy Labels by Sample Selection and Replay
    Luo, Yiwei
    Jiang, Min
    2023 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, IJCNN, 2023,
  • [44] RECALL: Replay-based Continual Learning in Semantic Segmentation
    Maracani, Andrea
    Michieli, Umberto
    Toldo, Marco
    Zanuttigh, Pietro
    2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021), 2021, : 7006 - 7015
  • [45] Saliency-Augmented Memory Completion for Continual Learning
    Bai, Guangji
    Ling, Chen
    Gao, Yuyang
    Zhao, Liang
    PROCEEDINGS OF THE 2023 SIAM INTERNATIONAL CONFERENCE ON DATA MINING, SDM, 2023, : 244 - 252
  • [46] An Energy Sampling Replay-Based Continual Learning Framework
    Zhang, Xingzhong
    Chuah, Joon Huang
    Loo, Chu Kiong
    Wermter, Stefan
    ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING-ICANN 2024, PT II, 2024, 15017 : 17 - 30
  • [47] Practical Recommendations for Replay-Based Continual Learning Methods
    Merlin, Gabriele
    Lomonaco, Vincenzo
    Cossu, Andrea
    Carta, Antonio
    Bacciu, Davide
    IMAGE ANALYSIS AND PROCESSING, ICIAP 2022 WORKSHOPS, PT II, 2022, 13374 : 548 - 559
  • [48] Replay-Based Continual Learning for Test Case Prioritization
    Fariha, Asma
    Azim, Akramul
    Liscano, Ramiro
    2024 IEEE INTERNATIONAL CONFERENCE ON SOFTWARE TESTING, VERIFICATION AND VALIDATION WORKSHOPS, ICSTW 2024, 2024, : 106 - 107
  • [49] CONTINUAL LEARNING WITH FOUNDATION MODELS: AN EMPIRICAL STUDY OF LATENT REPLAY
    Ostapenko, Oleksiy
    Lesort, Timothee
    Rodriguez, Pau
    Arefin, Md Rifat
    Douillard, Arthur
    Rish, Irina
    Charlin, Laurent
    CONFERENCE ON LIFELONG LEARNING AGENTS, VOL 199, 2022, 199
  • [50] Combining replay and LoRA for continual learning in natural language understanding
    Borhanifard, Zeinab
    Faili, Heshaam
    Yaghoobzadeh, Yadollah
    COMPUTER SPEECH AND LANGUAGE, 2025, 90