Self-BSR: Self-supervised Image Denoising and Destriping Based on Blind-Spot Regularization
We propose Self-BSR, a self-supervised image denoising and destriping method based on blind-spot regularization. This approach reformulates the overall denoising and destriping problem as a modeling task involving two spatially correlated signals: the image and the stripe.
Apr 9, 2025
Physics-guided infrared spatiotemporal noise modeling based on hybrid neural representation
We propose an infrared spatiotemporal noise modeling framework (IRSTN) based on hybrid neural representation, which leverages unpaired video data to simulate real-world noise.
Oct 23, 2024
Frequency-Aware Degradation Modeling for Real-World Thermal Image Super-Resolution
To enhance the SR performance for real-world thermal images, we propose an unsupervised SR framework that integrates degradation modeling with corresponding SR reconstruction.
Jan 16, 2024