Reinforcement Learning-Based Sequential Parameter Tuning for Image Signal Processing
IEEE Transactions on Pattern Analysis and Machine Intelligence · 2026-04 · 48 (4) · 4114-4131
DOI: 10.1109/tpami.2025.3642837 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026-04. Verfasst von Xinyu Sun und 7 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 66 Literaturangaben. Gefördert von: Beijing Natural Science - Xiaomi Innovation Joint Foundation, National Natural Science Foundation of China, Beijing Natural Science Foundation.
Abstract (Originalsprache)
Hardware image signal processing (ISP) transforms RAW inputs into high-quality RGB images through a series of processing modules, each with numerous tunable parameters. Traditionally, these parameters are manually tuned by imaging experts, a time-consuming and subjective process. Recent deep learning approaches predict ISP parameters, but often treat the process as a black box and overlook the intrinsic relationships among ISP modules. To address these fundamental issues, we introduce a novel ISP parameter optimization model based on single-agent reinforcement learning (RL) (i.e., SARL-ISP), formulating the hardware ISP parameter tuning as a sequential optimization problem. During the optimization process, the agent updates ISP parameter tuning strategies for different tasks through interaction with the environment. In order to explore the influence of the sequential structure of hardware ISP modules and the coupling relationships among ISP parameters on the tuning process, we further propose a sequential ISP framework based on collaborative multi-agent RL (i.e., MARL-ISP). Specifically, the serialized parameter tuning module (SPTM) realistically simulates the process of manual prediction and module pipeline. Additionally, the feature selection module (FSM) facilitates the transmission and fusion of agent features, thereby selecting more appropriate feature inputs for downstream tasks. Extensive experiments across various tasks (e.g., object detection, instance segmentation) validate the effectiveness and efficiency of our models. Even with minimal training data, our models also outperform current state-of-the-art methods in both quantitative metrics and qualitative evaluations.
Förderung
- Beijing Natural Science - Xiaomi Innovation Joint Foundation
- National Natural Science Foundation of China
- Beijing Natural Science Foundation
- Key Research and Development Program of Xinjiang Uyghur Autonomous Region
Zitieren
APA
Sun, X., Li, B., Lang, C., Zhao, Z., Wang, J., Xiong, W., Hu, W., & Cheng, L. (2026). Reinforcement Learning-Based Sequential Parameter Tuning for Image Signal Processing. IEEE Transactions on Pattern Analysis and Machine Intelligence, 48(4), 4114-4131. https://doi.org/10.1109/tpami.2025.3642837
Vancouver
Sun X, Li B, Lang C, Zhao Z, Wang J, Xiong W, et al. Reinforcement Learning-Based Sequential Parameter Tuning for Image Signal Processing. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2026;48(4):4114-4131. doi:10.1109/tpami.2025.3642837
BibTeX
@article{Sun2026,
author = {Sun, Xinyu and Li, Bing and Lang, Congyan and Zhao, Zhikun and Wang, Juan and Xiong, Weihua and Hu, Weiming and Cheng, Long},
title = {Reinforcement Learning-Based Sequential Parameter Tuning for Image Signal Processing},
journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
year = {2026},
volume = {48},
number = {4},
pages = {4114--4131},
doi = {10.1109/tpami.2025.3642837}
}
EUVIMED