Wei Yang

Groundtruth of the MPII test is held by the MPII team for fair comparison. For getting evaluation results on the MPII test set, you need to prepare your results as...

You can find the code to generate the JSON file here: https://github.com/bearpaw/pytorch-pose/blob/master/miscs/gen_lsp.m Actually, it is borrowed from Convolutional Pose Machine: https://github.com/shihenw/convolutional-pose-machines-release/blob/a345bc5c760073792b6b5ec29d3bde01557a5a9b/training/genJSON.m I didn't dig into the equations. But it should...

Another thing of CPM is that, it uses additional training data from LSP and LSP extended to train the model for the MPII dataset. In this case, we want to...

Hi @ooeygui , thank you very much for your response. Yes I think a low resolution model is very useful for robotics application (like you mentioned for collision, positioning and...

Got it, thanks! Is that the best model publicly available (just to make sure)?

Can we use the alpha in `colorrgba` message? https://github.com/microsoft/Azure_Kinect_ROS_Driver/blob/melodic/src/k4a_ros_device.cpp#L780

Please follow the instructions [here](https://github.com/bearpaw/pytorch-pose/blob/master/TRAINING.md#mpii-human-pose-dataset) to download the json file. You can also generate json file by running https://github.com/bearpaw/pytorch-pose/blob/master/miscs/gen_mpii.m

Please check https://github.com/bearpaw/pytorch-pose/issues/53

@achigeor Sorry I didn't check the correctness of this part. Did you get correct results or not?

There's no "human detection" here. The single-person human pose estimation benchmarks such as MPII/LSP/FLIC provide the location and the scale of the target person. So we assume the detection has...