[Pedestron] Generalizable Pedestrian Detection: The Elephant In The Room. @ CVPR2021

Cascade-RCNN+DeepSort MOTDT Trackor++

YOLOv7 training. Generates a head-only dataset in YOLO format. The labels included in the CrowdHuman dataset are Head and FullBody, but ignore FullBody.

[CVPR2022] "Progressive End-to-End Object Detection in Crowded Scenes" on Deformable-DETR.

This repo contains a script to convert the CrowdHuman dataset annotations to COCO format and a `dataset Class` for reading data.

YOLOv8 trained on CrowdHuman dataset and supports ONNX Runtime Inference for Image, Video and Webcam

crowdhuman dataset for darknet

Three Factor Authentication System, Add an additional layer of security to the old 2FA using Facial Biometric Scan.