Raul Balmez

VesperMUSE: Multi-Modal Fusion Vision Transformer for Image Semantic Segmentation Under Diverse Light Conditions

V VesperMUSE is introduced, a unified multi-modal framework designed to achieve robust image semantic segmentation across diverse illumination conditions and successfully recovers missing visual cues, providing a flexible, state-of-the-art (SOTA) solution for machine perception in extreme and unstructured real-world en...