MSR-VTT

Clone from “friedrichor/MSR-VTT”. MSRVTT contains 10K video clips and 200K captions. We adopt the standard 1K-A split protocol, which was introduced in JSFusion and has since become the de facto benchmark split in the Text-Video Retrieval field. Train: train 7k: 7,010 videos, 140,200 captions train 9k: 9,000 videos, 180,000 captions Test: test 1k: 1,000 videos, 1,000 captions 🌟 Citation @inproceedings{xu2016msrvtt, title={Msr-vtt: A large video description dataset… See the full description on the dataset page:

Source: Hugging Face Hub (VLM2Vec/MSR-VTT). Metadata imported from the dataset’s Hub tags.