Open Arbor334 opened 1 month ago
Hi @Arbor334, frame selection is done here, which calls modeling/sampler.py
to sample the frames.
Frame aggregation is done here in the similarity matrix computation step.
thanks!
gumbel_softmax_top_k(logits, top_k, tau, hard=False, reduce_sum=True, maskvector=True, logit_tau=1):
Is this function from modeling/gumbel_softmax.py
used to select top-k frames? and actions = gumbel_softmax_top_k(prob, int(self.top_k), tau, True, reduce_sum=False)
What does the return value actions mean? select_frames or not ? :smiley:
Great project, I looked at your
model.py
andtrain.py
files Can you pinpoint the core code for frame selection and frame aggregation in the file?:happy: