Search papers, labs, and topics across Lattice.
The paper introduces Regional Prior attention Transformer for infrared image Super-Resolution (RPT-SR) to address the inefficiency of general SR models in exploiting spatial priors in fixed-viewpoint infrared imaging. RPT-SR uses a dual-token framework, fusing learnable regional prior tokens (scene layout) with local tokens (frame content), to dynamically modulate local reconstruction via attention. Experiments demonstrate state-of-the-art performance across LWIR and SWIR datasets, showcasing the model's broad applicability.
By explicitly encoding scene layout priors into the attention mechanism, RPT-SR achieves state-of-the-art infrared image super-resolution, outperforming general-purpose transformers that waste capacity on redundant learning.
General-purpose super-resolution models, particularly Vision Transformers, have achieved remarkable success but exhibit fundamental inefficiencies in common infrared imaging scenarios like surveillance and autonomous driving, which operate from fixed or nearly-static viewpoints. These models fail to exploit the strong, persistent spatial priors inherent in such scenes, leading to redundant learning and suboptimal performance. To address this, we propose the Regional Prior attention Transformer for infrared image Super-Resolution (RPT-SR), a novel architecture that explicitly encodes scene layout information into the attention mechanism. Our core contribution is a dual-token framework that fuses (1) learnable, regional prior tokens, which act as a persistent memory for the scene's global structure, with (2) local tokens that capture the frame-specific content of the current input. By utilizing these tokens into an attention, our model allows the priors to dynamically modulate the local reconstruction process. Extensive experiments validate our approach. While most prior works focus on a single infrared band, we demonstrate the broad applicability and versatility of RPT-SR by establishing new state-of-the-art performance across diverse datasets covering both Long-Wave (LWIR) and Short-Wave (SWIR) spectra
