Lighted-SAM: Lightening Open-World SAM for Low-Light Segmentation
IEEE Transactions on Image Processing · 2026 · 35 · 6572-6584
DOI: 10.1109/tip.2026.3702382 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Image Processing, 2026. Verfasst von Yuhan Jia und 3 weiteren Autor:innen. Die Arbeit nennt 50 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Natural Science Foundation of Sichuan Province, Frontier Cross Innovation Team Project of Southwest Jiaotong University.
Abstract (Originalsprache)
Segment Anything Model (SAM) has achieved impressive segmentation performance in an open-world setting. However, SAM relies heavily on high-quality input images and usually struggles in low-light conditions. This is mainly caused by the pre-training dataset, SA-1B, in which low-light samples constitute a relatively small fraction of the data. This lack of presence leads to a noticeable weakness when SAM is applied in real-world dark environments. With the motivation of improving SAM's performance under low-light conditions while retaining its strong zero-shot capability, this work proposes an alignment stage between the pre-training stage and testing stage. Unlike existing low-light studies that mainly focus on task-specific and close-set settings, our work further emphasizes pursuing the segmentation ability under low-light conditions for open-world models. To this end, we construct DarkSeg58K, a realistic and diverse dataset, which serves as the alignment dataset to support this stage. We further introduce Lighted-SAM as the lightweight repair strategy to fix SAM's performance in low-light conditions. Different from existing methods focusing on introducing spectral adapters into the model design and training this model end-to-end, Lighted-SAM introduces the Spectral Information Resonance (SIR) mechanism to harmoniously integrate the spectral enhancement module into SAM, which is usually kept frozen due to its large-scale parameters. Based on our lightweight repairing strategy, Lighted-SAM can improve SAM's ability in low-light conditions while preserving its zero-shot ability. Experiments on different benchmarks validate the superiority of our approach. Code is available at: https://github.com/Jaaaahan/LightedSAM.
Förderung
- National Natural Science Foundation of China
- Natural Science Foundation of Sichuan Province
- Frontier Cross Innovation Team Project of Southwest Jiaotong University
- Fundamental Research Funds for the Central Universities
Zitieren
APA
Jia, Y., Duan, L., Li, W., & Lv, F. (2026). Lighted-SAM: Lightening Open-World SAM for Low-Light Segmentation. IEEE Transactions on Image Processing, 35, 6572-6584. https://doi.org/10.1109/tip.2026.3702382
Vancouver
Jia Y, Duan L, Li W, Lv F. Lighted-SAM: Lightening Open-World SAM for Low-Light Segmentation. IEEE Transactions on Image Processing. 2026;35:6572-6584. doi:10.1109/tip.2026.3702382
BibTeX
@article{Jia2026,
author = {Jia, Yuhan and Duan, Lixin and Li, Wen and Lv, Fengmao},
title = {Lighted-SAM: Lightening Open-World SAM for Low-Light Segmentation},
journal = {IEEE Transactions on Image Processing},
year = {2026},
volume = {35},
pages = {6572--6584},
doi = {10.1109/tip.2026.3702382}
}
EUVIMED