Aurora Series: AuroraCap - a wchai Collection

wchai 's Collections

Aurora Series: AuroraCap

STEVE

Aurora Series: AuroraCap

updated Oct 26

Efficient, Performant Video Detailed Captioning and a New Benchmark

AuroraCap: Efficient, Performant Video Detailed Captioning and a New Benchmark

Paper • 2410.03051 • Published Oct 4 • 4
wchai/AuroraCap-7B-VID-xtuner

Video-Text-to-Text • Updated Oct 7 • 72 • 3
wchai/AuroraCap-7B-IMG-xtuner

Image-Text-to-Text • Updated Oct 7 • 33 • 2
wchai/Video-Detailed-Caption

Viewer • Updated Oct 7 • 1.03k • 4.04k • 4

Note The VDC benchmark contains 1,027 videos with captions averaging over 500 words.
wchai/lmms_VDC_test

Viewer • Updated Oct 19 • 5.14k • 662 • 1

Note VDC benchmark in lmms-eval format.
wchai/AuroraCap-trainset

Preview • Updated Oct 13 • 937 • 7

Note over 20M image and video data collection for AuroraCap training with vicuna and llama-3 pre-tokenize.
wchai/AuroraCap-recaption

Viewer • Updated Oct 7 • 22.4k • 75 • 5

Note video data recaptioned by AuroraCap.