RGB Thermal

RGB-thermal (RGBT) research focuses on integrating visible and thermal imagery to improve computer vision tasks, overcoming limitations of each modality alone. Current efforts concentrate on developing effective data fusion techniques, often employing neural networks like Neural Radiance Fields (NeRF) and convolutional neural networks (CNNs) with asymmetric architectures to handle the inherent differences between RGB and thermal data, and exploring novel loss functions for synthetic data generation. This interdisciplinary field is significant for advancing applications in diverse areas such as autonomous driving, surveillance, and medical imaging, where robust performance in challenging lighting conditions is crucial.