About this document
Cross-Supervised Object Detection Framework by nahianibnasad is a document available to read on EtoBox.
This document presents a novel approach called cross-supervised object detection (CSOD) that enhances weakly supervised object detection by leveraging knowledge from fully labeled base categories to improve localization of novel objects using only class labels. The proposed framework integrates a detection head trained on instance-level annotations with a recognition head trained on image-level annotations, along with a spatial correlation module to bridge the gap between detection and recognition. Experime
- Author
- nahianibnasad
- Language
- EN