MCPcopy Create free account

hub / github.com/cavalleria/cavaface / functions

Functions528 in github.com/cavalleria/cavaface

⨍ Functions 528◇ Types & classes 126↳ Endpoints 4
↓ 1 callersMethod_load_faces_lists
(self)
evaluation/eval_megaface.py:84
↓ 1 callersFunction_load_json_result_file
(json_file)
evaluation/utils/io.py:85
↓ 1 callersFunction_load_model
(args)
evaluation/main.py:29
↓ 1 callersMethod_load_model
(self)
evaluation/infer/citrus_base_infer.py:276
↓ 1 callersFunction_main
(args)
evaluation/main.py:72
↓ 1 callersMethod_make_branches
(self, num_branches, block, num_blocks, num_channels)
backbone/hrnet.py:503
↓ 1 callersMethod_make_dataset
(self, root_dir, class_to_idx)
dataset/datasets.py:33
↓ 1 callersMethod_make_fuse_layers
(self)
backbone/hrnet.py:510
↓ 1 callersMethod_make_head
(self, pre_stage_channels)
backbone/hrnet.py:672
↓ 1 callersMethod_make_one_branch
(self, branch_index, block, num_blocks, num_channels, stride=1)
backbone/hrnet.py:467
↓ 1 callersFunction_merge_results
(args, eval_set, net_scale_type, test_type=None)
evaluation/evaluate_service.py:132
↓ 1 callersMethod_projection
(self, p, grad, perturb, delta, wd_ratio, eps)
optimizer/optimizer.py:456
↓ 1 callersMethod_projection
(self, p, grad, perturb, delta, wd_ratio, eps)
optimizer/optimizer.py:577
↓ 1 callersMethod_read_embedding_base
(self, file_list, emb_outdir, img_root)
evaluation/eval_ijbc.py:216
↓ 1 callersFunctionaugment_list
()
dataset/randaugment.py:152
↓ 1 callersFunctioncalculate_roc
(thresholds, embeddings1, embeddings2, actual_issame, nrof_folds=10, pca = 0)
evaluation/utils/io.py:315
↓ 1 callersFunctioncalculate_roc
( thresholds, embeddings1, embeddings2, actual_issame, nrof_folds=10, pca=0 )
util/verification.py:38
↓ 1 callersFunctioncalculate_val
(thresholds, embeddings1, embeddings2, actual_issame, far_target, nrof_folds=10)
evaluation/utils/io.py:379
↓ 1 callersMethodcollect_tensor
Collect all tensor objects tracked by python NOTICE: - the buffers for backward which is implemented in C++ are not t
evaluation/utils/pytorch_memlab/mem_reporter.py:49
↓ 1 callersFunctionconv_1x1_bn
(inp, oup)
backbone/mobilenetv3.py:80
↓ 1 callersFunctionconv_3x3_bn
(inp, oup, stride)
backbone/mobilenetv3.py:72
↓ 1 callersFunctioncount_model_flops
(model, input_res=[112, 112], multiply_adds=True)
util/flops_counter.py:35
↓ 1 callersFunctioncount_model_params
(model)
util/flops_counter.py:27
↓ 1 callersFunctioncutmix_data
(input, target, gpu, cutmix_prob=0.5, alpha=1.0)
dataset/utils.py:187
↓ 1 callersMethoddisable
(self)
evaluation/utils/pytorch_memlab/line_profiler.py:95
↓ 1 callersMethoddo_flip
(self, data)
evaluation/infer/citrus_base_infer.py:125
↓ 1 callersFunctiondwconv3x3_block
3x3 depthwise version of the standard convolution block.
backbone/common.py:790
↓ 1 callersFunctionestimate_unit_mem
(model, shape, ctx)
evaluation/utils/pytorch_tool.py:17
↓ 1 callersFunctionevaluate
(embeddings, actual_issame, nrof_folds=10, pca=0)
util/verification.py:202
↓ 1 callersMethodextract_embedding
(self, net)
evaluation/eval_ijbc.py:190
↓ 1 callersFunctionflops_to_string
(flops, units="GFLOPS", precision=2)
util/flops_counter.py:7
↓ 1 callersFunctionfoo
(net)
util/flops_counter.py:103
↓ 1 callersMethodforward_features
(self, x)
backbone/hrnet.py:820
↓ 1 callersFunctionget_activation_layer
Create activation layer from string/function. Parameters: ---------- activation : function, or str, or nn.Module Activation
backbone/common.py:98
↓ 1 callersFunctionget_blocks
(num_layers)
backbone/resnet_irse.py:31
↓ 1 callersFunctionget_infer
(args)
evaluation/infer/__init__.py:3
↓ 1 callersMethodget_lr
(self)
optimizer/lr_scheduler.py:90
↓ 1 callersFunctionget_min_free_gpu_mem
(gpuid)
evaluation/utils/io.py:439
↓ 1 callersMethodget_num_inchannels
(self)
backbone/hrnet.py:580
↓ 1 callersFunctionget_optimizer
(optimizer_name, params, lr, momentum)
optimizer/optimizer.py:645
↓ 1 callersMethodget_params
(img, scale, ratio)
dataset/utils.py:131
↓ 1 callersMethodget_stats
Get the memory stat of tensors and then release them As a memory profiler, we cannot hold the reference to any tensors, which causes
evaluation/utils/pytorch_memlab/mem_reporter.py:64
↓ 1 callersFunctionget_time
()
util/utils.py:15
↓ 1 callersMethodimage2template_feature
(self, img_feats=None, templates=None, medias=None)
evaluation/eval_ijbc.py:98
↓ 1 callersMethodinit_weights
(self)
backbone/hrnet.py:812
↓ 1 callersFunctionload_bin
(path, image_size=(112, 112))
util/verification.py:282
↓ 1 callersFunctionmain
()
train.py:29
↓ 1 callersMethodmetric
(self)
evaluation/eval_ijbc.py:240
↓ 1 callersFunctionmixup_criterion
(criterion, pred, y_a, y_b, lam)
dataset/utils.py:75
↓ 1 callersFunctionmixup_data
Returns mixed inputs, pairs of targets, and lambda
dataset/utils.py:60
↓ 1 callersFunctionparams_to_string
(params_num, units="M", precision=2)
util/flops_counter.py:18
↓ 1 callersMethodparse_results_into_file
(self)
evaluation/eval_ijbc.py:301
↓ 1 callersFunctionpre_conv3x3_block
3x3 version of the pre-activated convolution block.
backbone/common.py:980
↓ 1 callersMethodprint_stats
(self, verbose=False, target_device=None)
evaluation/utils/pytorch_memlab/mem_reporter.py:134
↓ 1 callersMethodprocs_params
(self)
head/metrics_parallel.py:37
↓ 1 callersFunctionrand_bbox
(size, lam)
dataset/utils.py:168
↓ 1 callersFunctionread_samples_from_record
(root_dir, record_dir, Train)
dataset/datasets.py:64
↓ 1 callersMethodread_template_media_list
(self, path)
evaluation/eval_ijbc.py:73
↓ 1 callersMethodread_template_pair_list
(self, path)
evaluation/eval_ijbc.py:79
↓ 1 callersMethodremove_noisy_embedding
(self)
evaluation/eval_megaface.py:181
↓ 1 callersMethodreport
Interface for end-users to directly print the memory usage args: - verbose: flag to show tensor.storage reuse information
evaluation/utils/pytorch_memlab/mem_reporter.py:178
↓ 1 callersMethodreset
(self)
util/verification.py:229
↓ 1 callersMethodreset_parameters
(self)
head/metrics.py:432
↓ 1 callersMethodreset_params
(self)
head/metrics_parallel.py:44
↓ 1 callersMethodreset_steps
(self, start_step, nr_steps, start_value=0, stop_value=None)
backbone/common.py:475
↓ 1 callersMethodrestore
Restore the tensors into original CUDA devices
evaluation/utils/pytorch_memlab/courtesy.py:32
↓ 1 callersMethodrun
(self)
evaluation/evaluate_service.py:93
↓ 1 callersFunctionseparate_bn_paras
(modules)
util/utils.py:47
↓ 1 callersFunctionseparate_irse_bn_paras
(modules)
util/utils.py:25
↓ 1 callersFunctionset_seed
(seed)
util/utils.py:7
↓ 1 callersMethodslice_params
(self)
head/metrics_parallel.py:47
↓ 1 callersFunctionsmooth_one_hot
(true_labels, classes, smoothing=0.0)
loss/loss.py:37
↓ 1 callersFunctionstart_eval_process
(eval_info)
evaluation/main.py:41
↓ 1 callersFunctionstart_merge_process
(merge_info)
evaluation/main.py:58
↓ 1 callersMethodstep
(self)
backbone/common.py:538
↓ 1 callersFunctiontest
(data_set, backbone, batch_size, nfolds=10)
util/verification.py:310
↓ 1 callersFunctionver_test
(ver_dataset, backbone)
util/verification.py:271
↓ 1 callersMethodverification
( self, template_norm_feats=None, unique_templates=None, p1=None, p2=None )
evaluation/eval_ijbc.py:133
↓ 1 callersMethodyield_memory
Transfer all the CUDA tensors into CPU memory
evaluation/utils/pytorch_memlab/courtesy.py:14
↓ 1 callersMethodzero_grad
(self)
optimizer/optimizer.py:51
FunctionAttentionNet_IRSE_56
Constructs a AttentionNet_IRSE_56 model.
backbone/resattnet.py:383
FunctionAttentionNet_IRSE_92
Constructs a AttentionNet_IRSE_92 model.
backbone/resattnet.py:397
FunctionAttentionNet_IR_56
Constructs a AttentionNet_IR_56 model.
backbone/resattnet.py:376
FunctionAttentionNet_IR_92
Constructs a AttentionNet_IR_92 model.
backbone/resattnet.py:390
FunctionAutoContrast
(img, _)
dataset/randaugment.py:62
FunctionBrightness
(img, v)
dataset/randaugment.py:108
FunctionColor
(img, v)
dataset/randaugment.py:103
FunctionContrast
(img, v)
dataset/randaugment.py:98
FunctionCutout
(img, v)
dataset/randaugment.py:118
FunctionDIR_FAR
Closed/Open-set Identification. A general case of Cummulative Match Characteristic (CMC) where thresholding is allowed for open-set
util/metrics.py:151
FunctionEqualize
(img, _)
dataset/randaugment.py:70
FunctionFlip
(img, _)
dataset/randaugment.py:74
FunctionHRNet_W18
(input_size)
backbone/hrnet.py:882
FunctionHRNet_W18_small
(input_size)
backbone/hrnet.py:872
FunctionHRNet_W18_small_v2
(input_size)
backbone/hrnet.py:877
FunctionHRNet_W30
(input_size)
backbone/hrnet.py:887
FunctionHRNet_W32
(input_size)
backbone/hrnet.py:892
FunctionHRNet_W40
(input_size)
backbone/hrnet.py:897
FunctionHRNet_W44
(input_size)
backbone/hrnet.py:902
FunctionHRNet_W48
(input_size)
backbone/hrnet.py:907
← previousnext →101–200 of 528, ranked by callers