[Question]Parsing good poses and poses from LSTM module ?

[timtensor]

Hi , I am trying to extend theLSTM moduleand was wondering if it could be used to make a simple parser ,to differentiate Good posture and Bad Posture.

What i have is lets say Ref Video which lets say is a sequence of situps done correctly and test video where some situups are wrong (ex 3/10 ) . I am extracting the normalized keypoints in json format x, y co-ordinates and confidence score (Coco -posenet format )per frame for both reference and test video .

I want to extract the those 3 bad situps. Is it possible to do so ?? If so could you provide me the guidance.

[gohjiayi]

I think that it is possible by training the model with data for good and bad situps. It will be hard though because of the similarity between good/bad posture and the fact that pose estimation flickers a lot when joints are not detected.

I'm looking to implement posture correction & rep counting but not too sure on how to do so. Any help is appreciated, looking forward.

[timtensor]

@gohjiayi sorry for the late reply. I kind of discontinued after struggling on distinguishing on goodbad poses. I was looking to explore more the action recognition side of things. I saw you had made some progress, perhaps i can DM you ?

[gohjiayi]

@timtensor Sure, reach me via contact methods linked on my profile. Cheers!