Efficient Continual Learning through Frequency Decomposition and Integration
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866912298928439296 |
|---|---|
| author | Liu, Ruiqi Diao, Boyu Huang, Libo Liu, Hangda Yang, Chuanguang An, Zhulin Xu, Yongjun |
| author_facet | Liu, Ruiqi Diao, Boyu Huang, Libo Liu, Hangda Yang, Chuanguang An, Zhulin Xu, Yongjun |
| contents | Continual learning (CL) aims to learn new tasks while retaining past knowledge, addressing the challenge of forgetting during task adaptation. Rehearsal-based methods, which replay previous samples, effectively mitigate forgetting. However, research on enhancing the efficiency of these methods, especially in resource-constrained environments, remains limited, hindering their application in real-world systems with dynamic data streams. The human perceptual system processes visual scenes through complementary frequency channels: low-frequency signals capture holistic cues, while high-frequency components convey structural details vital for fine-grained discrimination. Inspired by this, we propose the Frequency Decomposition and Integration Network (FDINet), a novel framework that decomposes and integrates information across frequencies. FDINet designs two lightweight networks to independently process low- and high-frequency components of images. When integrated with rehearsal-based methods, this frequency-aware design effectively enhances cross-task generalization through low-frequency information, preserves class-specific details using high-frequency information, and facilitates efficient training due to its lightweight architecture. Experiments demonstrate that FDINet reduces backbone parameters by 78%, improves accuracy by up to 7.49% over state-of-the-art (SOTA) methods, and decreases peak memory usage by up to 80%. Additionally, on edge devices, FDINet accelerates training by up to 5$\times$. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_22175 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Efficient Continual Learning through Frequency Decomposition and Integration Liu, Ruiqi Diao, Boyu Huang, Libo Liu, Hangda Yang, Chuanguang An, Zhulin Xu, Yongjun Computer Vision and Pattern Recognition Continual learning (CL) aims to learn new tasks while retaining past knowledge, addressing the challenge of forgetting during task adaptation. Rehearsal-based methods, which replay previous samples, effectively mitigate forgetting. However, research on enhancing the efficiency of these methods, especially in resource-constrained environments, remains limited, hindering their application in real-world systems with dynamic data streams. The human perceptual system processes visual scenes through complementary frequency channels: low-frequency signals capture holistic cues, while high-frequency components convey structural details vital for fine-grained discrimination. Inspired by this, we propose the Frequency Decomposition and Integration Network (FDINet), a novel framework that decomposes and integrates information across frequencies. FDINet designs two lightweight networks to independently process low- and high-frequency components of images. When integrated with rehearsal-based methods, this frequency-aware design effectively enhances cross-task generalization through low-frequency information, preserves class-specific details using high-frequency information, and facilitates efficient training due to its lightweight architecture. Experiments demonstrate that FDINet reduces backbone parameters by 78%, improves accuracy by up to 7.49% over state-of-the-art (SOTA) methods, and decreases peak memory usage by up to 80%. Additionally, on edge devices, FDINet accelerates training by up to 5$\times$. |
| title | Efficient Continual Learning through Frequency Decomposition and Integration |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.22175 |