qvhighlights
qvhighlights-videos
QVHighlights Videos
All 25124 videos from the QVHighlights benchmark (train + val + test splits).
Total size: 135.3 GB.
Layout
Files are sharded into subdirectories by the first character of the filename
(HuggingFace caps each directory at 10,000 files):
<first-char>/<youtube-id>_<start>_<end>.mp4
Source
Lei et al., "QVHighlights: Detecting Moments and Highlights in Videos via Natural
Language Queries" (NeurIPS 2021).
Original archive:… See the full description on the dataset page: https://huggingface.co/datasets/ayushsdev/qvhighlights-videos.qvhighlights-1fps
QVHighlights 1fps — Preprocessed Frames
Preprocessed version of the QVHighlights dataset for temporal video grounding.
Videos are extracted at 1fps, resized to 384×384 JPEG, ready for training without any video I/O at runtime.
Contents
File
Description
annotations_train.jsonl
7445 train annotations
annotations_val.jsonl
1550 val annotations
frames folder
Train frames batch 0000–1000
frames_000000_001000.tar
Train frames batch 0000–1000… See the full description on the dataset page: https://huggingface.co/datasets/shaunmarvell/qvhighlights-1fps.QVHighlights_preprocessedTrying to add dataset files
QVHighlightsqvhighlights-valqvhighlights-test