rohit901/VANE-Bench

[NAACL'25] Contains code and documentation for our VANE-Bench paper.

38
/ 100
Emerging

This project offers a specialized benchmark for evaluating how well advanced AI models (Video-LMMs) can spot unusual or inconsistent events in video footage. It takes in video clips, both synthetically generated and from real-world surveillance, and generates question-answer pairs about anomalies. This tool is for AI researchers, particularly those working on computer vision and multi-modal AI, to test and improve their anomaly detection models.

No commits in the last 6 months.

Use this if you are a researcher or developer working with Video-LMMs and need a rigorous way to test their ability to detect subtle, unexpected events or inconsistencies in video content.

Not ideal if you are looking for a plug-and-play solution for real-time anomaly detection in production environments, as this is a research benchmark for model evaluation.

AI model evaluation video anomaly detection computer vision research multi-modal AI AI safety
Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 14 / 25

How are scores calculated?

Stars

23

Forks

4

Language

Python

License

MIT

Last pushed

Aug 19, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/rohit901/VANE-Bench"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.