EUVIMEDEUVIMED entdecken
Crossref

HotMoE: Exploring Sparse Mixture-of-Experts for Hyperspectral Object Tracking

Wenfang Sun, Yuedong Tan, Jingyuan Li, Shuwei Hou, Xiaobo Li, Yingzhao Shao, Zhe Wang, Beibei Song

IEEE Transactions on Multimedia · 2025 · 27 · 4072-4083

DOI: 10.1109/tmm.2025.3535339 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Multimedia, 2025. Verfasst von Wenfang Sun und 7 weiteren Autor:innen. Die Arbeit wurde bisher 27-mal zitiert (Crossref) und nennt 68 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Key Research and Development Projects of Shaanxi Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Hyperspectral videos contain richer spectral and physical features than RGB videos and thus have greater potential for use in object tracking. The mainstream hyperspectral object tracking approach involves the integration of multiple RGB-based video tracking models. Although ensembles of multiple models can effectively utilize spectral information and improve tracker performance, this approach has high computational complexity, making it difficult to meet the real-time requirements of video object tracking. To bridge the gap, we propose a new hyperspectral object tracking framework (HotMoE) based on Mixture-of-Experts (MoE). HotMoE leverages a divide-and-conquer strategy, where only a subset of expert models is computed for each input, reducing computational complexity while maintaining performance. In this paper, we first design a splitter to group multiple spectral bands into multiple false-color images based on spectral correlations. Then, we design a hyperspectral MoE router that can adaptively learn to aggregate spectral image feature information and route it to suitable experts. Different experts can handle various scenarios, and HotMoE effectively utilizes the capabilities of different experts to obtain better overall performance. Compared with previous state-of-the-art hyperspectral object tracking networks, our model has significantly reduced inference time and performs well, with a processing speed of 43.7 FPS and an AUC of 0.704 with the HOT2022 dataset.

Förderung

  • National Natural Science Foundation of China
  • Key Research and Development Projects of Shaanxi Province

Zitieren

APA

Sun, W., Tan, Y., Li, J., Hou, S., Li, X., Shao, Y., Wang, Z., & Song, B. (2025). HotMoE: Exploring Sparse Mixture-of-Experts for Hyperspectral Object Tracking. IEEE Transactions on Multimedia, 27, 4072-4083. https://doi.org/10.1109/tmm.2025.3535339

Vancouver

Sun W, Tan Y, Li J, Hou S, Li X, Shao Y, et al. HotMoE: Exploring Sparse Mixture-of-Experts for Hyperspectral Object Tracking. IEEE Transactions on Multimedia. 2025;27:4072-4083. doi:10.1109/tmm.2025.3535339

BibTeX
@article{Sun2025,
  author = {Sun, Wenfang and Tan, Yuedong and Li, Jingyuan and Hou, Shuwei and Li, Xiaobo and Shao, Yingzhao and Wang, Zhe and Song, Beibei},
  title = {HotMoE: Exploring Sparse Mixture-of-Experts for Hyperspectral Object Tracking},
  journal = {IEEE Transactions on Multimedia},
  year = {2025},
  volume = {27},
  pages = {4072--4083},
  doi = {10.1109/tmm.2025.3535339}
}
Kostenlos starten →