CLIP-Joint-Detect: End-to-End Joint Training of Object Detectors with Contrastive Vision-Language Supervision

Behnam Raoufi, Hossein Sharify, Mohamad Mahdee Ramezanee +2

Conventional object detectors rely on cross-entropy classification, which can be vulnerable to class imbalance and label noise. We propose CLIP-Joint-Detect, a simple and detector-…