Thanks for sharing the code for your paper. you stated that "For RGB modality, we set the number of segments to 8 and the
number of frames sampled in each segment to 6". In Table 5, where you compare the TSM with different inputs, are the number of RGB frames used for the network as follows:
model : number of RGB frames
TSM : 48 (8*6)
TSM + Cropping (ours) : 48
TSM + STMEM (ours): 8
TSM + Cropping + STMEM (ours): 8
DSCNet (ours) : 8?
Thanks for sharing the code for your paper. you stated that "For RGB modality, we set the number of segments to 8 and the
number of frames sampled in each segment to 6". In Table 5, where you compare the TSM with different inputs, are the number of RGB frames used for the network as follows:
model : number of RGB frames
TSM : 48 (8*6)
TSM + Cropping (ours) : 48
TSM + STMEM (ours): 8
TSM + Cropping + STMEM (ours): 8
DSCNet (ours) : 8?