EUVIMEDEUVIMED entdecken
Crossref

Cross-Modal Knowledge Distillation for Oriented Object Detection in Modality Missing Visible-Infrared Images

Yifan Xi, Ting Lu, Jiacheng Lu, Xudong Kang, Shutao Li

IEEE Transactions on Circuits and Systems for Video Technology · 2026-07 · 36 (7) · 9363-9377

DOI: 10.1109/tcsvt.2026.3670458 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Circuits and Systems for Video Technology, 2026-07. Verfasst von Yifan Xi und 4 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 76 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Outstanding Young Talents Program of Changsha, Science and Technology Innovation Program of Hunan Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Visible and infrared light images reflect object characteristics in different aspects, which has attracted much attention for object detection in recent years. Nevertheless, the existing multimodal detection networks may fail in the absence of modality. In order to address this problem, a new cross-modal knowledge distillation network (CMKD-net) is proposed for oriented object detection in visible and infrared images. In brief, a teacher-student (T-S) learning network is constructed, where the T-network aims to learn a discriminative feature representation from multimodal images and then guides the S-network training with incomplete modality. Here, multi-dimensional feature distillation (MDFD) and inter-instance relation distillation (IIRD) are designed for cross-modal knowledge propagation. Specifically, the MDFD considers pushing the T-S networks to learn a similar data distribution and feature representation through channel-spatial dimensional feature consistency constraints. The IIRD contributes to retaining the relation structure between individual targets in multimodal images via inter-instance relation modeling and similarity distance measurement. Moreover, to avoid the bias of feature extraction caused by discrete quantization in traditional pooling operations, a rotation-adaptive RoI Pooling (RA-RoI Pooling) is introduced by calculating the continuous double integral within each bin of oriented objects. Ablation experiments and comparison experiments on the VEDAI and DroneVehicle datasets can demonstrate the effectiveness of the proposed CMKD-net.

Förderung

  • National Natural Science Foundation of China
  • Outstanding Young Talents Program of Changsha
  • Science and Technology Innovation Program of Hunan Province

Zitieren

APA

Xi, Y., Lu, T., Lu, J., Kang, X., & Li, S. (2026). Cross-Modal Knowledge Distillation for Oriented Object Detection in Modality Missing Visible-Infrared Images. IEEE Transactions on Circuits and Systems for Video Technology, 36(7), 9363-9377. https://doi.org/10.1109/tcsvt.2026.3670458

Vancouver

Xi Y, Lu T, Lu J, Kang X, Li S. Cross-Modal Knowledge Distillation for Oriented Object Detection in Modality Missing Visible-Infrared Images. IEEE Transactions on Circuits and Systems for Video Technology. 2026;36(7):9363-9377. doi:10.1109/tcsvt.2026.3670458

BibTeX
@article{Xi2026,
  author = {Xi, Yifan and Lu, Ting and Lu, Jiacheng and Kang, Xudong and Li, Shutao},
  title = {Cross-Modal Knowledge Distillation for Oriented Object Detection in Modality Missing Visible-Infrared Images},
  journal = {IEEE Transactions on Circuits and Systems for Video Technology},
  year = {2026},
  volume = {36},
  number = {7},
  pages = {9363--9377},
  doi = {10.1109/tcsvt.2026.3670458}
}
Kostenlos starten →