I am using a pretrained model for pose estimation with Pytorch: https://github.com/microsoft/human-pose-estimation.pytorch
I am using the model to detect persons and their 17 keypoints in an image but I am not sure how I should convert the output tensor to a format of pixel coordinates for each keypoint with a confidence score for each person.
This is the code for the prediction:
model.eval()
img_t = ToTensor()(img).unsqueeze(0)
out = model(img_t)
out.shape
The output shape is [1, 16, 64, 56]. I am a bit lost. Any help would be appreciated.
