CS-REG-NET: A Self-Supervised Visual-State-Space based Architecture for Cross-Spectral Registration of Thermal and Optical Imagery CS-REG-NET: Termal ve Optik G r nt lerde apraz-Spektral aki stirma i in G rsel-Durum Uzayi Tabanli zg zetimli grenmeli Mimari
33rd IEEE Conference on Signal Processing and Communications Applications, SIU 2025, İstanbul, Türkiye, 25 - 28 Haziran 2025, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/siu66497.2025.11112294
- Basıldığı Şehir: İstanbul
- Basıldığı Ülke: Türkiye
- Anahtar Kelimeler: cross-spectral image matching, homography estimation, self-supervised learning, VMamba
- İstanbul Medipol Üniversitesi Adresli: Evet
Özet
Modern deep models for multispectral image matching typically rely on large, supervised datasets, which can be prohibitively expensive. To overcome this challenge, we introduce CS-REG-NET, a self-supervised, detector-based framework that requires no external labels. Instead, it uses RIFT2 detector to generate pseudo-ground-truth keypoints. A VMamba encoder, pre-trained on a segmentation task, processes image pairs, while two output heads learn feature heatmaps and descriptors. CSREG-NET significantly outperforms existing methods, delivering superior keypoint detection and homography estimation. This real-time framework thus provides a robust, extensible solution for multispectral image matching.