Human Pose Estimation

This repository provides implementation for Human Pose Estimation that predicts the location of various human keypoints (joints and landmarks) such as elbows, knees, neck, shoulder, hips, chest etc.

Requirements

• OpenCV
• OpenPose

Usage

• To download the pre-Trained models, run sh get_model.sh command.
• To run pre-Trained model, trained on MPII Human Pose dataset, on any image, run the following command:- python3 human_pose_estimation.py --input <Image Name> --proto human_pose_proto/mpi/pose_deploy_linevec_faster_4_stages.prototxt --model human_pose_proto/mpi/pose_iter_160000.caffemodel --dataset MPI.
• To run pre-Trained model, trained on COCO dataset, on any image, run the following command:- python3 human_pose_estimation.py --input <Image Name> --proto human_pose_proto/coco/deploy_coco.prototxt --model human_pose_proto/coco/pose_iter_440000.caffemodel --dataset COCO.
• To run pre-Trained model, trained on Body_25 dataset, on any image, run the following command:- python3 human_pose_estimation.py --input <Image Name> --proto human_pose_proto/body_25/body_25_deploy.prototxt --model human_pose_proto/body_25/pose_iter_584000.caffemodel.

Results

Following are a few human poses predicted by the model:-

Test Image	Generated Keypoint Skeleton