MCPcopy Create free account

hub / github.com/MCG-NJU/VideoMAE / functions

Functions287 in github.com/MCG-NJU/VideoMAE

⨍ Functions 287◇ Types & classes 48
↓ 83 callersFunctionprint
(*args, **kwargs)
utils.py:212
↓ 35 callersMethodupdate
(self, **kwargs)
utils.py:88
↓ 11 callersFunction_cfg
(url='', **kwargs)
modeling_finetune.py:11
↓ 9 callersMethodload_state_dict
(self, state_dict)
utils.py:366
↓ 7 callersFunction_check_args_tf
(kwargs)
rand_augment.py:58
↓ 6 callersFunctioncutmix_bbox_and_lam
Generate bbox and apply lambda correction.
mixup.py:77
↓ 6 callersMethodloadvideo_decord
Load video content using Decord
kinetics.py:222
↓ 6 callersMethodloadvideo_decord
Load video content using Decord
ssv2.py:217
↓ 6 callersMethodstate_dict
(self)
utils.py:363
↓ 5 callersMethod__init__
(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.)
modeling_finetune.py:36
↓ 5 callersFunction_randomly_negate
With 50% prob, negate the value
rand_augment.py:195
↓ 5 callersFunctiontrunc_normal_
(tensor, mean=0., std=1.)
modeling_pretrain.py:14
↓ 4 callersFunction_get_param_spatial_crop
Given scale, ratio, height and width, return sampled coordinates of the videos.
video_transforms.py:499
↓ 4 callersMethod_params_per_elem
(self, batch_size)
mixup.py:121
↓ 4 callersMethodadd_meter
(self, name, meter)
utils.py:117
↓ 4 callersMethodlog_every
(self, iterable, print_freq, header=None)
utils.py:120
↓ 4 callersMethodmax
(self)
utils.py:67
↓ 4 callersMethodset_step
(self, step=None)
utils.py:172
↓ 4 callersMethodsynchronize_between_processes
(self)
utils.py:113
↓ 3 callersMethod__len__
(self)
kinetics.py:274
↓ 3 callersMethod__len__
(self)
ssv2.py:265
↓ 3 callersFunction_pil_interp
(method)
video_transforms.py:33
↓ 3 callersFunctionblend
Blend two images with a given weight alpha. Args: images1 (tensor): the first images to be blended, the dimension is `num
video_transforms.py:276
↓ 3 callersFunctionbuild_dataset
(is_train, test_mode, args)
datasets.py:57
↓ 3 callersFunctionget_sinusoid_encoding_table
Sinusoid position encoding table
modeling_finetune.py:160
↓ 3 callersFunctionis_dist_avail_and_initialized
()
utils.py:220
↓ 3 callersMethodupdate
(self, head='scalar', step=None, **kwargs)
utils.py:178
↓ 2 callersMethod__init__
(self, img_size=224, patch_size=16, encoder_in_chans=3,
modeling_pretrain.py:181
↓ 2 callersMethod_aug_frame
( self, buffer, args, )
kinetics.py:160
↓ 2 callersMethod_aug_frame
( self, buffer, args, )
ssv2.py:155
↓ 2 callersMethod_erase
(self, img, chan, img_h, img_w, dtype)
random_erasing.py:80
↓ 2 callersFunction_get_pixels
( per_pixel, rand_color, patch_size, dtype=torch.float32, device="cuda" )
random_erasing.py:11
↓ 2 callersFunction_load_checkpoint_for_ema
Workaround for ModelEma._load_checkpoint to accept an already-loaded object
utils.py:195
↓ 2 callersMethod_params_per_batch
(self)
mixup.py:141
↓ 2 callersFunctionconvert_img
Converts (H, W, C) numpy.ndarray to (C, W, H) format
volume_transforms.py:6
↓ 2 callersFunctioncreate_optimizer
(args, model, get_num_layer=None, get_layer_scale=None, filter_bias_and_bn=True, skip_list=None)
optim_factory.py:91
↓ 2 callersFunctioncrop_boxes
Peform crop on the bounding boxes given the offsets. Args: boxes (ndarray or None): bounding boxes to peform crop. The dimension
video_transforms.py:101
↓ 2 callersFunctionfinal_test
(data_loader, model, device, file)
engine_for_finetuning.py:180
↓ 2 callersMethodflush
(self)
utils.py:187
↓ 2 callersFunctionget_parameter_groups
(model, weight_decay=1e-5, skip_list=(), get_num_layer=None, get_layer_scale=None)
optim_factory.py:49
↓ 2 callersFunctionget_resize_sizes
(im_h, im_w, size)
functional.py:67
↓ 2 callersFunctiongrayscale
Get the grayscale for the input images. The channels of images should be in order BGR. Args: images (tensor): the input images fo
video_transforms.py:292
↓ 2 callersFunctionmerge
(eval_path, num_tasks)
engine_for_finetuning.py:233
↓ 2 callersFunctionmixup_target
(target, num_classes, lam=1., smoothing=0.0, device='cuda')
mixup.py:22
↓ 2 callersMethodno_weight_decay
(self)
modeling_finetune.py:251
↓ 2 callersFunctionone_hot
(x, num_classes, on_value=1., off_value=0., device='cuda')
mixup.py:17
↓ 2 callersFunctionrand_augment_transform
RandAugment: Practical automated data augmentation... - https://arxiv.org/abs/1909.13719 Create a RandAugment transform :param config_st
rand_augment.py:481
↓ 2 callersFunctiontrain_class_batch
(model, samples, target, criterion)
engine_for_finetuning.py:13
↓ 2 callersMethodupdate
(self, value, n=1)
utils.py:34
↓ 1 callersMethod_erase_cube
( self, img, batch_start, batch_size, chan, img_h, img
random_erasing.py:109
↓ 1 callersFunction_interpolation
(kwargs)
rand_augment.py:50
↓ 1 callersFunction_is_tensor_clip
(clip)
functional.py:8
↓ 1 callersMethod_make_dataset
(self, directory, setting)
kinetics.py:498
↓ 1 callersMethod_mix_batch
(self, x)
mixup.py:196
↓ 1 callersMethod_mix_batch_collate
(self, output, batch)
mixup.py:280
↓ 1 callersMethod_mix_elem
(self, x)
mixup.py:159
↓ 1 callersMethod_mix_elem_collate
(self, output, batch, half=False)
mixup.py:227
↓ 1 callersMethod_mix_pair
(self, x)
mixup.py:176
↓ 1 callersMethod_mix_pair_collate
(self, output, batch)
mixup.py:252
↓ 1 callersFunction_posterize_level_to_arg
(level, _hparams)
rand_augment.py:242
↓ 1 callersMethod_sample_crop_size
(self, im_size)
transforms.py:111
↓ 1 callersMethod_sample_fix_offset
(self, image_w, image_h, crop_w, crop_h)
transforms.py:135
↓ 1 callersMethod_sample_train_indices
(self, num_frames)
kinetics.py:515
↓ 1 callersFunction_select_rand_weights
(weight_idx=0, transforms=None)
rand_augment.py:444
↓ 1 callersFunction_solarize_level_to_arg
(level, _hparams)
rand_augment.py:263
↓ 1 callersMethod_video_TSN_decord_batch_loader
(self, directory, video_reader, duration, indices, skip_offsets)
kinetics.py:538
↓ 1 callersFunctionbrightness_jitter
Perfrom brightness jittering on the input images. The channels of images should be in order BGR. Args: var (float): jitter ratio
video_transforms.py:349
↓ 1 callersFunctionbuild_pretraining_dataset
(args)
datasets.py:38
↓ 1 callersFunctioncontrast_jitter
Perfrom contrast jittering on the input images. The channels of images should be in order BGR. Args: var (float): jitter ratio fo
video_transforms.py:368
↓ 1 callersMethodfill_fix_offset
(more_fix_crop, image_w, image_h, crop_w, crop_h)
transforms.py:140
↓ 1 callersMethodforward_features
(self, x)
modeling_finetune.py:261
↓ 1 callersMethodforward_features
(self, x, mask)
modeling_pretrain.py:88
↓ 1 callersFunctionget_args
()
run_mae_pretraining.py:19
↓ 1 callersFunctionget_args
()
run_class_finetuning.py:27
↓ 1 callersFunctionget_args
()
run_videomae_vis.py:48
↓ 1 callersFunctionget_grad_norm_
(parameters, norm_type: float = 2.0)
utils.py:370
↓ 1 callersFunctionget_loss_scale_for_deepspeed
(model)
engine_for_finetuning.py:19
↓ 1 callersFunctionget_model
(args)
run_mae_pretraining.py:121
↓ 1 callersFunctionget_model
(args)
run_videomae_vis.py:75
↓ 1 callersFunctionget_num_layer_for_vit
(var_name, num_max_layer)
optim_factory.py:24
↓ 1 callersMethodget_num_layers
(self)
modeling_finetune.py:247
↓ 1 callersMethodget_params
Get parameters for ``crop`` for a random sized crop. Args: img (PIL Image): Image to be cropped. scale (tuple): range
video_transforms.py:727
↓ 1 callersMethodget_params
(self, brightness, contrast, saturation, hue)
video_transforms.py:1191
↓ 1 callersFunctionget_position_angle_vec
(position)
modeling_finetune.py:163
↓ 1 callersFunctionget_rank
()
utils.py:234
↓ 1 callersFunctionget_world_size
()
utils.py:228
↓ 1 callersFunctionis_main_process
()
utils.py:240
↓ 1 callersFunctionload
(module, prefix='')
utils.py:302
↓ 1 callersFunctionmain
(args)
run_mae_pretraining.py:134
↓ 1 callersFunctionmain
(args, ds_init)
run_class_finetuning.py:208
↓ 1 callersFunctionmain
(args)
run_videomae_vis.py:88
↓ 1 callersFunctionrand_augment_ops
(magnitude=10, hparams=None, transforms=None)
rand_augment.py:453
↓ 1 callersFunctionrand_bbox
Standard CutMix bounding-box Generates a random square bbox based on lambda value. This impl includes support for enforcing a border margin a
mixup.py:30
↓ 1 callersFunctionrand_bbox_minmax
Min-Max CutMix bounding-box Inspired by Darknet cutmix impl, generates a random rectangular bbox based on min/max percent values applied to e
mixup.py:54
↓ 1 callersFunctionsaturation_jitter
Perfrom saturation jittering on the input images. The channels of images should be in order BGR. Args: var (float): jitter ratio
video_transforms.py:388
↓ 1 callersFunctionsave_on_master
(*args, **kwargs)
utils.py:244
↓ 1 callersFunctionsetup_for_distributed
This function disables printing when not in master process
utils.py:205
↓ 1 callersFunctionspatial_sampling
Perform spatial sampling on the given video frames. If spatial_idx is -1, perform random scale, random crop, and random flip on the given
kinetics.py:281
↓ 1 callersFunctionspatial_sampling
Perform spatial sampling on the given video frames. If spatial_idx is -1, perform random scale, random crop, and random flip on the given
ssv2.py:272
↓ 1 callersFunctiontensor_normalize
Normalize a given tensor by subtracting the mean and dividing the std. Args: tensor (tensor): tensor to normalize. mean (tens
kinetics.py:355
next →1–100 of 287, ranked by callers