Real time Fight Detection Based on 2D Pose Estimation and RNN Action Recognition
-
Updated
May 20, 2020 - C++
Real time Fight Detection Based on 2D Pose Estimation and RNN Action Recognition
Fight Detection From Surveillance Cameras by fine-tuning a PyTorch Pretrained Model
Fight detection using pose estimation, limb angle calculation and LSTM
AI-powered real-time Fight and Non-Fight Detection System using VideoMAE, R3D-18, YOLO26n, and OpenVINO for intelligent video surveillance.
Real-time fight detection on surveillance cameras with YOLOv8 classification, trained on RWF-2000: motion gating, 5-frame decision window, cooldown, evidence clips and bagging ensemble support.
To associate your repository with the fight-detection topic, visit your repo's landing page and select "manage topics."