Use YOLO Keypoints for Secondary GIE (LSTM Classifier)

[al0nkr]

Hi, I wanted to use a custom LSTM Pose Classifier model on top of the yolo generated keypoints, I want yoloV8 keypoints and apply post processing on them to fit into my classifier model. I tried implementing a custom parser by modifying the nvdsparsepose_Yolo.cpp file, however I could not understand how can i access the Primary GIE's output (yolo keypoints) and pass them to another Secondary GIE.

The expected output format for the yolo keypoints should be of the shape (1 x 17 x 2) which, after some post processing should be converted to a tensor of shape (1 x42). I need the converted tensor as input to my LSTM Classifier model, please let me know if you need some other reference files to go about this issue.

[marcoslucianops]

The DeepStream doesn't support non image input. You must to create a custom logic for it.

[al0nkr]

So is there no way to pass keypoints / model outputs of primary GIE to the secondary GIE Natively? If you were to implement this custom logic, can you provide me some rough ideas to go about creating it? As in, how can i store the key point values and parse it with custom logic to take it as a non image input? Thanks!

[zhendong-yu]

The problem is resolved after the version is specified as 8.0.100；
$ pip install ultralytics==8.0.100