Feature drift is caused by the dynamic coupling of target features and degradation factors, which reduce underwater detector performance. We redefine feature drift as the instability of target ...

Machine learning-based object detection is important in a wide range of fields and applications, where it identifies and locates objects within images or video frames. It plays an important role in a ...

Results of Class-agnostic Object Detection of MViTS including our proposed Multiscale Attention ViT with Late fusion (MAVL) model, applications, and exploratory analysis. Class-agnostic Object ...