MCPcopy Create free account

hub / github.com/Lu-Feng/DHE-VPR / functions

Functions131 in github.com/Lu-Feng/DHE-VPR

⨍ Functions 131◇ Types & classes 25
↓ 8 callersFunctioncct_7
(arch, pretrained, progress, aggregation=None, *args, **kwargs)
model/cct/cct.py:144
↓ 5 callersMethod__init__
(self, p=3, eps=1e-6)
network.py:27
↓ 5 callersMethod__init__
(self, dim, num_heads=8, attention_dropout=0.1, projection_dropout=0.1)
model/cct/transformers.py:12
↓ 5 callersFunction_cct
(arch, pretrained, progress, num_layers, num_heads, mlp_ratio, embedding_dim, kernel_size=3,
model/cct/cct.py:102
↓ 5 callersFunctionpath_to_pil_img
(path)
datasets_ws.py:24
↓ 4 callersFunctioncct_6
(arch, pretrained, progress, aggregation=None, *args, **kwargs)
model/cct/cct.py:139
↓ 3 callersFunctioncct_14
(arch, pretrained, progress, aggregation=None, *args, **kwargs)
model/cct/cct.py:149
↓ 3 callersMethodget_positives
(self)
datasets_ws.py:139
↓ 2 callersMethod__init__
(self, args, datasets_folder="dataset", dataset_folder="pitts30k/images/train")
datasets_ws.py:52
↓ 2 callersMethod__iter__
(self)
commons.py:42
↓ 2 callersFunctioncct_2
(arch, pretrained, progress, aggregation=None, *args, **kwargs)
model/cct/cct.py:129
↓ 2 callersFunctioncct_4
(arch, pretrained, progress, aggregation=None, *args, **kwargs)
model/cct/cct.py:134
↓ 2 callersMethodcompute_cache
Compute the cache containing features of images, which is used to find best positive and hardest negatives.
datasets_ws.py:245
↓ 2 callersMethodget_best_positive_index
(self, args, model, query_index, cache, cache_a, query_features)
datasets_ws.py:274
↓ 2 callersMethodget_hardest_negatives_indexes
(self, args, cache, query_features, neg_samples)
datasets_ws.py:283
↓ 2 callersMethodget_query_features
(self, query_index, cache)
datasets_ws.py:266
↓ 2 callersFunctionmatch_batch_tensor
fm1: (l,D) fm2: (N,l,D) mask1: (l) mask2: (N,l)
homography_project.py:22
↓ 2 callersMethodregression
(self, similarity_matrix)
network.py:174
↓ 2 callersMethodsimilarity
(self, tensor_img_1, tensor_img_2)
network.py:168
↓ 1 callersMethod__getitem__
(self, index)
datasets_ws.py:57
↓ 1 callersMethod__init__
(self, kernel_size, stride, padding, pooling_kernel_size=3, pooling_stride=2
model/cct/tokenizer.py:7
↓ 1 callersMethod__len__
(self)
datasets_ws.py:59
↓ 1 callersMethod_test_query_transform
Transform query image according to self.test_method.
datasets_ws.py:114
↓ 1 callersFunctioncct_14_7x2_384
(pretrained=False, progress=False, img_size=384, positional_embedding='learnable', num_clas
model/cct/cct.py:342
↓ 1 callersFunctioncompute_recalls
Computes the recalls. Parameters ---------- predictions : np.array of int, containing the first 20 predictions for each query,
util.py:72
↓ 1 callersFunctioncompute_similarity
(features_a, features_b)
network.py:110
↓ 1 callersMethodcompute_triplets
(self, args, model)
datasets_ws.py:237
↓ 1 callersMethodcompute_triplets_full
(self, args, model)
datasets_ws.py:293
↓ 1 callersMethodcompute_triplets_partial
(self, args, model)
datasets_ws.py:322
↓ 1 callersFunctiondrop_path
Obtained from: github.com:rwightman/pytorch-image-models Drop paths (Stochastic Depth) per sample (when applied in main path of residual bloc
model/cct/stochastic_depth.py:8
↓ 1 callersMethodforward
(self, x)
model/cct/tokenizer.py:43
↓ 1 callersMethodforward
(self, x, mask=None)
model/cct/tokenizer.py:99
↓ 1 callersMethodforward_mask
(self, mask)
model/cct/tokenizer.py:82
↓ 1 callersMethodforward_mask
(self, mask)
model/cct/embedder.py:18
↓ 1 callersFunctiongem
(x, p=3, eps=1e-6)
network.py:36
↓ 1 callersFunctionget_keypoints
(img_size)
homography_project.py:9
↓ 1 callersFunctionpe_check
(model, state_dict, pe_key='classifier.positional_emb')
model/cct/helpers.py:26
↓ 1 callersFunctionresize_pos_embed
(posemb, posemb_new, num_tokens=1)
model/cct/helpers.py:6
↓ 1 callersMethodsequence_length
(self, n_channels=3, height=224, width=224)
model/cct/tokenizer.py:40
↓ 1 callersMethodsinusoidal_embedding
(n_channels, dim)
model/cct/transformers.py:223
↓ 1 callersMethodsinusoidal_embedding
(n_channels, dim, padding_idx=False)
model/cct/transformers.py:329
↓ 1 callersFunctionto_cuda
Move to cuda all items of the list.
train_dhe.py:24
Method__getitem__
(self, index)
datasets_ws.py:103
Method__getitem__
(self, index)
datasets_ws.py:216
Method__getitem__
(self, index)
datasets_ws.py:373
Method__getitem__
(self, index)
datasets_ws.py:396
Method__getitem__
(self, index)
dataset_geoloc.py:57
Method__getitem__
(self, index)
dataset_qp.py:52
Method__init__
(self)
network.py:40
Method__init__
(self, dim=1)
network.py:47
Method__init__
(self, args)
network.py:58
Method__init__
(self, output_dim=16)
network.py:87
Method__init__
(self, features_extractor, homography_regression)
network.py:128
Method__init__
(self, args, datasets_folder="datasets", dataset_name="pitts30k", split="train")
datasets_ws.py:66
Method__init__
(self, args, datasets_folder="datasets", dataset_name="pitts30k", split="train", negs_num_per_query=10)
datasets_ws.py:150
Method__init__
(self, shape, dtype=np.float32)
datasets_ws.py:365
Method__init__
(self, shape, dtype=np.float32)
datasets_ws.py:386
Method__init__
(self, *args, **kwargs)
commons.py:39
Method__init__
Parameters ---------- datasets_folder : str, path of the folder with the datasets. dataset_name : str, name
dataset_geoloc.py:13
Method__init__
Dataset used to compute pairs of query-positive. Parameters ---------- model : nn.Module, used to compute the query-
dataset_qp.py:9
Method__init__
(self, drop_prob=None)
model/cct/stochastic_depth.py:34
Method__init__
(self, img_size=224, embedding_dim=768, n_input_channels=3,
model/cct/cct.py:33
Method__init__
(self, kernel_size, stride, padding, pooling_kernel_size=3, pooling_stride=2
model/cct/tokenizer.py:53
Method__init__
(self, dim, num_heads=8, attention_dropout=0.1, projection_dropout=0.1)
model/cct/transformers.py:39
Method__init__
(self, d_model, nhead, dim_feedforward=2048, dropout=0.1, attention_dropout=0.1, drop_path_ra
model/cct/transformers.py:78
Method__init__
(self, d_model, nhead, dim_feedforward=2048, dropout=0.1, attention_dropout=0.1, drop_path_ra
model/cct/transformers.py:108
Method__init__
(self, seq_pool=True, embedding_dim=768, num_layers=12,
model/cct/transformers.py:134
Method__init__
(self, seq_pool=True, embedding_dim=768, num_layers=12,
model/cct/transformers.py:232
Method__init__
(self, word_embedding_dim=300, vocab_size=100000, padding_i
model/cct/embedder.py:5
Method__len__
(self)
datasets_ws.py:135
Method__len__
(self)
datasets_ws.py:230
Method__len__
(self)
dataset_geoloc.py:62
Method__len__
(self)
dataset_qp.py:61
Method__next__
(self)
commons.py:44
Method__repr__
(self)
network.py:33
Method__repr__
(self)
datasets_ws.py:137
Method__repr__
(self)
dataset_geoloc.py:65
Method__setitem__
(self, indexes, vals)
datasets_ws.py:369
Method__setitem__
(self, indexes, vals)
datasets_ws.py:390
Functioncct_14_7x2_224
(pretrained=False, progress=False, img_size=224, positional_embedding='learnable', num_clas
model/cct/cct.py:331
Functioncct_14_7x2_384_fl
(pretrained=False, progress=False, img_size=384, positional_embedding='learnable', num_c
model/cct/cct.py:353
Functioncct_2_3x2_32
(pretrained=False, progress=False, img_size=32, positional_embedding='learnable', num_classes
model/cct/cct.py:155
Functioncct_2_3x2_32_sine
(pretrained=False, progress=False, img_size=32, positional_embedding='sine', num_classes
model/cct/cct.py:166
Functioncct_4_3x2_32
(pretrained=False, progress=False, img_size=32, positional_embedding='learnable', num_classes
model/cct/cct.py:177
Functioncct_4_3x2_32_sine
(pretrained=False, progress=False, img_size=32, positional_embedding='sine', num_classes
model/cct/cct.py:188
Functioncct_6_3x1_32
(pretrained=False, progress=False, img_size=32, positional_embedding='learnable', num_classes
model/cct/cct.py:199
Functioncct_6_3x1_32_sine
(pretrained=False, progress=False, img_size=32, positional_embedding='sine', num_classes
model/cct/cct.py:210
Functioncct_6_3x2_32
(pretrained=False, progress=False, img_size=32, positional_embedding='learnable', num_classes
model/cct/cct.py:221
Functioncct_6_3x2_32_sine
(pretrained=False, progress=False, img_size=32, positional_embedding='sine', num_classes
model/cct/cct.py:232
Functioncct_7_3x1_32
(pretrained=False, progress=False, img_size=32, positional_embedding='learnable', num_classes
model/cct/cct.py:243
Functioncct_7_3x1_32_c100
(pretrained=False, progress=False, img_size=32, positional_embedding='learnable', num_cl
model/cct/cct.py:265
Functioncct_7_3x1_32_sine
(pretrained=False, progress=False, img_size=32, positional_embedding='sine', num_classes
model/cct/cct.py:254
Functioncct_7_3x1_32_sine_c100
(pretrained=False, progress=False, img_size=32, positional_embedding='sine', num_cl
model/cct/cct.py:276
Functioncct_7_3x2_32
(pretrained=False, progress=False, img_size=32, positional_embedding='learnable', num_classes
model/cct/cct.py:287
Functioncct_7_3x2_32_sine
(pretrained=False, progress=False, img_size=32, positional_embedding='sine', num_classes
model/cct/cct.py:298
Functioncct_7_7x2_224
(pretrained=False, progress=False, img_size=224, positional_embedding='learnable', num_class
model/cct/cct.py:309
Functioncct_7_7x2_224_sine
(pretrained=False, progress=False, img_size=224, positional_embedding='sine', num_class
model/cct/cct.py:320
Functioncollate_fn
Creates mini-batch tensors from the list of tuples (images, triplets_local_indexes, triplets_global_indexes). triplets_local_indexes
datasets_ws.py:28
Functioncompute_features
Compute the features of all images within the geoloc_dataset. Parameters ---------- geoloc_dataset : dataset_geoloc.GeolocDataset, w
util.py:9
Functioncompute_score
(model, tensor_img_1, tensor_img_2, weights=None)
homography_project.py:105
next →1–100 of 131, ranked by callers