A Graph Attention Spatio-temporal Convolutional Network for 3D Human Pose Estimation in Video

Benchmark Model Rank Results
3d-human-pose-estimation-on-humaneva-iGAST#7Mean Reconstruction Error (mm): 21.2