Incorporating Dense Knowledge Alignment into Unified Multimodal Representation Models

Benchmark Model Rank Results
zero-shot-cross-modal-retrieval-on-coco-2014DeKR-7B (zero-shot)–Image-to-text R@1: 73.3
zero-shot-video-retrieval-on-msr-vttDeKR-7B (zero-shot)–text-to-video R@1: 42.9
zero-shot-video-retrieval-on-msr-vtt-video-to-textDeKR-7B (zero-shot)–R@1: 43.7