TheShadow29/VidSitu
[CVPR21] Visual Semantic Role Labeling for Video Understanding (https://arxiv.org/abs/2104.00990)
VidSitu helps researchers analyze short video clips by identifying the verbs, semantic roles, and relationships between events within them. It takes 10-second movie clips as input and outputs detailed annotations of what is happening, who is doing it, and how different actions are connected, all at 2-second intervals. This project is for computer vision researchers and AI model developers working on understanding complex video content.
No commits in the last 6 months.
Use this if you are a computer vision researcher developing or evaluating models for understanding human actions and events in short video segments.
Not ideal if you need a plug-and-play solution for real-time video analysis or for analyzing very long-form video content without segmenting.
Stars
61
Forks
8
Language
Python
License
MIT
Category
Last pushed
Aug 17, 2021
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/TheShadow29/VidSitu"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
TheShadow29/awesome-grounding
awesome grounding: A curated list of research papers in visual grounding
microsoft/XPretrain
Multi-modality pre-training
TheShadow29/zsgnet-pytorch
Official implementation of ICCV19 oral paper Zero-Shot grounding of Objects from Natural...
zeyofu/BLINK_Benchmark
This repo contains evaluation code for the paper "BLINK: Multimodal Large Language Models Can...
gicheonkang/sglkt-visdial
🌈 PyTorch Implementation for EMNLP'21 Findings "Reasoning Visual Dialog with Sparse Graph...