Infrared image super-resolution with fewer parameters and faster processing
TIRMamba: A Thermal-Prior-Modulated State-Space Network for Sub-Million-Parameter Infrared Image Super-Resolution
Computer Vision and Pattern RecognitionMachine Learning
Summary
Infrared images are often blurry and need special tools to make them clearer, but existing tools are very large and slow for devices like drones or handheld cameras. The authors created a much smaller and faster network called TIRMamba that uses smart ways to focus on thermal image features to improve resolution. They trained it on both infrared and regular images to make it effective despite limited infrared data. Their model matches or beats larger existing methods while being much lighter and quicker, helping thermal imaging work better on small devices.
What this means in practice
- •For drone developers: Implement TIRMamba to enhance thermal image clarity on drones without heavy computational load or latency issues.
- •For handheld thermal camera manufacturers: Integrate TIRMamba to improve infrared image resolution on portable devices while maintaining fast processing and low power usage.$Commercial implications: Makes possible compact thermal cameras with high-resolution images suitable for commercial and industrial markets.
Authors
Chun-An Lin, Tsung-Jung Liu, Yen-Chieh Ouyang
Abstract
Infrared image super-resolution is currently led by Mamba-based networks with 26 to 37 million parameters, which are difficult to deploy on the airborne and handheld platforms where thermal imaging is most needed. This paper presents TIRMamba, a network with 896K to 910K parameters for single-channel thermal imagery. A Thermal Prior Highway computes gradient, local-contrast and spectral cues once at the input and, through one adapter per residual group, modulates a weight-tied bidirectional state-space trunk and gates its dual-scale detail branch; a tri-path reconstruction adds the learned residual to a bicubic radiometric baseline. Because the standard benchmark provides only 265 infrared training images and evaluates fusion products on full images, we train with a replay strategy: grayscale DIV2K pre-training followed by fine-tuning on 64-pixel patches drawn with equal probability from the infrared and natural corpora. At scale factor 4, TIRMamba matches the strongest protocol-trained methods on both official test sets with 29 to 40 times fewer parameters and 2.8 to 9.4 times lower latency; at scale factor 2 it gives the highest SSIM on both. A variant with prior-conditioned selectivity, TIRMamba-Rad, corrects a 3 dB raw-thermal failure of an intermediate size-invariant design and gives the best results at scale factor 4 on raw-thermal, unmanned-aerial-vehicle and independent-sensor test sets. Code and trained models will be released at https://github.com/julian135707/TIRMamba upon acceptance.