↓ 1 callersFunctioninterpolate_temporal_pos_embed temp_embed_old: (1, num_frames_old, 1, d) Returns: temp_embed_new: (1, num_frames_new, 1, d)
multi_modality/models/utils.py:56
↓ 1 callersFunctionmake_batch_data_sampler(
dataset, sampler, aspect_grouping, videos_per_batch, num_iters=None, start_iter=0, drop_last=False
)
single_modality/action_detection/alphaction/dataset/build.py:95
↓ 1 callersMethodmask(
self,
input_ids,
vocab_size,
device,
targets=None,
masked_in
multi_modality/models/criterions.py:275
↓ 1 callersFunctionseparate_batch_per_person :param things: [b, max_others, c, t, h, w] :return [n, max_others, c, t, h, w]
single_modality/action_detection/alphaction/modeling/roi_heads/action_head/IA_structure.py:368