Closed ashwinvaswani closed 4 years ago
So just wanted some clarification: Given a video, we need to segment the frames corresponding to the strokes(action) right? And it must be done for both players. So as you suggested, we need to develop a 3D model, as we have temporal dimension. Is there anything I'm missing or you have something else in your mind?
@Sharanranjit There's another issue on this. I'm closing this one. I'm currently working on that one, let me know if you want to contribute to it.
dataset extractor is complete. Work on action segmentation can be started