AFFS: Adaptive Fast Frequency Selection Algorithm for Deep Learning Feature Extraction
文献类型：期刊论文
| 作者 | Li, Xiaocan1,6; Xie, Kun1,6; He, Zilong1,6; Wen, Jigang2; Cao, Jiannong3; Zhang, Guangxing4; Xie, Gaogang5; Liang, Wei2 |
| 刊名 | IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
 |
| 出版日期 | 2026-02-01 |
| 卷号 | 38期号:2页码:842-856 |
| 关键词 | Feature extraction Discrete cosine transforms Adaptation models Training Data models Computational modeling Time-frequency analysis Deep learning Logic gates Image classification Discrete cosine transform feature extraction frequency components selection frequency domain |
| ISSN号 | 1041-4347 |
| DOI | 10.1109/TKDE.2025.3638836 |
| 英文摘要 | As deep learning (DL) continues to advance, effective feature extraction from large-scale data remains crucial for enhancing model performance. To leverage the advantages of the frequency domain, such as concentrated signal energy, prominent data features, and rich detailed characteristics, this paper proposes a novel frequency-domain feature extraction method. However, existing frequency component selection algorithms often struggle to adapt to diverse tasks, tend to yield only locally optimal solutions, and require prolonged processing times. To overcome these limitations, we introduce the Adaptive Fast Frequency Selection (AFFS) algorithm, which seamlessly integrates a frequency component selection factor layer into DL models to identify globally optimal frequency combinations suited to various downstream tasks. We further analyze the relationship between selected frequency components and model performance, providing theoretical guarantees regarding optimality, robustness, and generalization error bounds. Moreover, a fast selection procedure is developed to exploit the empirically observed rapid convergence of the selection-factor ranking, significantly accelerating the selection process. Extensive experiments on five datasets, ten DL models, and two subsequent tasks demonstrate that AFFS achieves superior performance: even when the input data size is reduced to only 10% of the original frequency features, model classification accuracy improves by approximately 1%, while the early stopping mechanism shortens the selection process by about 80%. |
| 资助项目 | Key Research and Development Program of Hunan Province[2025QK3003] ; National Natural Science Foundation of China[62025201] ; National Natural Science Foundation of China[62472159] ; National Natural Science Foundation of China[62472167] ; National Natural Science Foundation of China[62321166652] ; Science and Technology Innovation Program of Hunan Province[2025RC3081] |
| WOS研究方向 | Computer Science ; Engineering |
| 语种 | 英语 |
| WOS记录号 | WOS:001655693200021 |
| 出版者 | IEEE COMPUTER SOC |
| 源URL | [http://119.78.100.204/handle/2XEOYT63/42817]  |
| 专题 | 中国科学院计算技术研究所 |
| 通讯作者 | Xie, Kun |
| 作者单位 | 1.Hunan Univ, Key Lab Fus Comp Supercomp & Artificial Intelligen, Minist Educ, Changsha 410082, Peoples R China 2.Hunan Univ Sci & Technol, Sch Comp Sci & Engn, Xiangtan 411201, Peoples R China 3.Hong Kong Polytech Univ, Dept Comp, Hong Kong, Peoples R China 4.Chinese Acad Sci, Inst Comp Technol, Dept Network Technol Res Ctr, Beijing 100864, Peoples R China 5.Chinese Acad Sci, Comp Network Informat Ctr, Beijing 100864, Peoples R China 6.Hunan Univ, Coll Comp Sci & Elect Engn, Changsha 410006, Peoples R China |
| 推荐引用方式 GB/T 7714 | Li, Xiaocan,Xie, Kun,He, Zilong,et al. AFFS: Adaptive Fast Frequency Selection Algorithm for Deep Learning Feature Extraction[J]. IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING,2026,38(2):842-856. |
| APA | Li, Xiaocan.,Xie, Kun.,He, Zilong.,Wen, Jigang.,Cao, Jiannong.,...&Liang, Wei.(2026).AFFS: Adaptive Fast Frequency Selection Algorithm for Deep Learning Feature Extraction.IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING,38(2),842-856. |
| MLA | Li, Xiaocan,et al."AFFS: Adaptive Fast Frequency Selection Algorithm for Deep Learning Feature Extraction".IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING 38.2(2026):842-856. |
入库方式： OAI收割
来源：计算技术研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。