Search Results for author:
Found 2 papers, 0 papers with code
Leveraging Open-Vocabulary Diffusion to Camouflaged Instance Segmentation
Such cross-domain representations are desirable in segmenting camouflaged objects where visual cues are subtle to distinguish the objects from the background, especially in segmenting novel objects which are not seen in training.
CDN-MEDAL: Two-stage Density and Difference Approximation Framework for Motion Analysis
Background modeling and subtraction is a promising research area with a variety of applications for video surveillance.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
