Video Event Detection Challenge
Spot the crisis, detect and localise!

CHALLENGE

Real-world events like person violence and falls are inherently spatio-temporal. Yet traditional pipelines-relying on object detectors and handcrafted post-processing-fail in crowded settings and miss key temporal context.
This AI challenge aims to develop an edge-deployable, end-to-end deep learning solution for real-time spatio-temporal event localization on live video streams.

HIGHLIGHTS

  • Data: Video clips (3-5 seconds) from diverse scenarios in lab.
  • Goal: Spatio-temporal localisation – detect and localise event.
  • Tech Stack: Use PyTorch, OpenCV, Python. Bonus points for real-time efficiency.
  • Evaluation: Automated metrics + expert review. Leaderboard updates frequently.

Why Compete? 

Internship: Top 3 winners secure paid summer internships at Vehant Research Labs-an environment to dive into edge AI deployment, and to collaborate on safety tech: A pathway to full-time roles.

Interested In Participation?Please Click Here To Register

 

IMPORTANT DATES

Event Date
Registration Open December 17, 2025
Release of Training Dataset January 09, 2026
Release of Test Set Images January 16, 2026
Submission Start Date January 19, 2026
Submission End Date January 30, 2026
Winner Announcement February 13, 2026

AWARDS

Position Prize
Winner 30,000
Runner-up 25,000

CONTACT 

Organizers: Abhay Kumar, Rahul Nair, Sannidhya Rawat, Prakhar Gupta

For any query please contact: contest@vehant.com