Salvato in:
Dettagli Bibliografici
Autori principali: Li, Zhifei, Wang, Yiran, Xiong, Chenyi, Xia, Yujing, Hou, Xiaoju, Zhao, Yue, Zhang, Miao, Xiao, Kui, Yang, Bing
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2601.01926
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914233369755648
author Li, Zhifei
Wang, Yiran
Xiong, Chenyi
Xia, Yujing
Hou, Xiaoju
Zhao, Yue
Zhang, Miao
Xiao, Kui
Yang, Bing
author_facet Li, Zhifei
Wang, Yiran
Xiong, Chenyi
Xia, Yujing
Hou, Xiaoju
Zhao, Yue
Zhang, Miao
Xiao, Kui
Yang, Bing
contents Visual Question Answering (VQA) requires models to reason over multimodal information, combining visual and textual data. With the development of continual learning, significant progress has been made in retaining knowledge and adapting to new information in the VQA domain. However, current methods often struggle with balancing knowledge retention, adaptation, and robust feature representation. To address these challenges, we propose a novel framework with adaptive memory allocation and global noise filtering called MacVQA for visual question answering. MacVQA fuses visual and question information while filtering noise to ensure robust representations, and employs prototype-based memory allocation to optimize feature quality and memory usage. These designs enable MacVQA to balance knowledge acquisition, retention, and compositional generalization in continual VQA learning. Experiments on ten continual VQA tasks show that MacVQA outperforms existing baselines, achieving 43.38% average accuracy and 2.32% average forgetting on standard tasks, and 42.53% average accuracy and 3.60% average forgetting on novel composition tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2601_01926
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle MacVQA: Adaptive Memory Allocation and Global Noise Filtering for Continual Visual Question Answering
Li, Zhifei
Wang, Yiran
Xiong, Chenyi
Xia, Yujing
Hou, Xiaoju
Zhao, Yue
Zhang, Miao
Xiao, Kui
Yang, Bing
Computer Vision and Pattern Recognition
Visual Question Answering (VQA) requires models to reason over multimodal information, combining visual and textual data. With the development of continual learning, significant progress has been made in retaining knowledge and adapting to new information in the VQA domain. However, current methods often struggle with balancing knowledge retention, adaptation, and robust feature representation. To address these challenges, we propose a novel framework with adaptive memory allocation and global noise filtering called MacVQA for visual question answering. MacVQA fuses visual and question information while filtering noise to ensure robust representations, and employs prototype-based memory allocation to optimize feature quality and memory usage. These designs enable MacVQA to balance knowledge acquisition, retention, and compositional generalization in continual VQA learning. Experiments on ten continual VQA tasks show that MacVQA outperforms existing baselines, achieving 43.38% average accuracy and 2.32% average forgetting on standard tasks, and 42.53% average accuracy and 3.60% average forgetting on novel composition tasks.
title MacVQA: Adaptive Memory Allocation and Global Noise Filtering for Continual Visual Question Answering
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2601.01926