Spectral conditional diffusion improves hyperspectral target detection accuracy
SCDM-HTD: Spectral-Wise Conditional Diffusion Model for Hyperspectral Target Detection
Computer Vision and Pattern Recognition
Summary
Hyperspectral target detection finds specific objects in detailed remote sensing images, but current methods often overlook learning how the background and target differ in overall distribution, making detection less reliable. The authors introduce a new approach that treats detection as a task of generating a map distinguishing targets from backgrounds based on their spectral information. They use a lightweight neural network and a diffusion model to better separate and identify targets, speeding up the process while improving accuracy. Tests on several standard datasets show this method works better than previous ones.
What this means in practice
- •For remote sensing analysts: Generate more accurate maps highlighting specific targets in hyperspectral remote sensing data to support environmental monitoring and land use assessment.
- •For defense surveillance teams: Improve automatic detection of targets in complex backgrounds using hyperspectral imaging for enhanced situational awareness in security operations.
Authors
Dunbin Shen, Wenfeng Kong, Jianjun Liu, Xiaorui Ma, Hongyu Wang
Abstract
Hyperspectral target detection (HTD) aims to locate and recognize objects at a fine-grained scale, playing a pivotal role in remote sensing. Despite significant progress, current HTD methods primarily focus on feature comparison while neglecting distribution learning for background-target separation, which can limit detection robustness in practical applications. This paper addresses this issue by introducing a spectral-wise conditional diffusion model that reformulates HTD as a task of generating a detection map. Specifically, we treat background and target confidence scores as training samples and use the corresponding pixel spectra as guiding conditions to capture the distributional discrepancy between background and target through diffusion and denoising. To enhance practical applicability, we propose a lightweight conditional denoising network based on multilayer perceptron to map noisy confidence scores to their ideal counterparts and adopt a deterministic reverse non-Markov chain to skip sampling steps, thereby accelerating the generation process. Furthermore, we introduce a composite loss to improve background-target separation. This loss integrates constrained energy minimization loss to enhance target saliency and suppress the background by maximizing the target response while minimizing average output energy, alongside binary cross-entropy loss to classify background and target using pixel-wise labels. Experimental results on four benchmark datasets demonstrate the effectiveness of the proposed method, with improvements in both quantitative metrics and qualitative visual effects.