MCPcopy Create free account

hub / github.com/cvlab-epfl/perspective-flow-aggregation / functions

Functions115 in github.com/cvlab-epfl/perspective-flow-aggregation

↓ 6 callersMethod__init__
(self, hidden_dim=128, input_dim=192+128)
flow/core/update.py:16
↓ 5 callersMethod_make_layer
(self, dim, stride=1)
flow/core/extractor.py:242
↓ 4 callersMethod__init__
(self, input_dim=3, output_dim=128, norm_fn='batch', dropout=0.0)
flow/core/extractor.py:120
↓ 3 callersMethod_make_layer
(self, dim, stride=1)
flow/core/extractor.py:160
↓ 3 callersMethod_make_layer
(self, dim, stride=1)
flow/core/extractor.py:336
↓ 3 callersMethodstep
(self)
flow/scheduler.py:45
↓ 2 callersMethodclose
(self)
flow/train.py:142
↓ 2 callersFunctioncoords_grid
(batch, ht, wd)
flow/core/utils/utils.py:74
↓ 2 callersFunctionevaluate_pose_predictions
(predictions, class_number, meshes, mesh_diameters, symmetry_types={})
pose_utils.py:48
↓ 2 callersFunctionis_memory_enough
()
flow/core/datasets.py:117
↓ 2 callersFunctionload_json_cached
(json_path, mem_cache=None)
flow/core/datasets.py:133
↓ 2 callersFunctionprint_accuracy_per_class
(accuracy_adi_per_class, accuracy_auc_per_class, accuracy_rep_per_class)
pose_utils.py:193
↓ 2 callersMethodstep
(self, optimizer)
flow/train.py:48
↓ 2 callersMethodupdate
(self)
flow/train.py:50
↓ 1 callersFunctionAddJitter2Pose
(K, R, T, max_angle=10, max_pix_offset=10)
flow/core/datasets.py:268
↓ 1 callersFunctionFlowEstimation
(flowmodel, image1, image2)
refinement.py:82
↓ 1 callersFunctionGetFlowFromPoseAndDepth
(pose1, pose2, K, depth1)
flow/core/datasets.py:231
↓ 1 callersFunctionGetVertmapFromDepth
(K, pose1, depth1)
flow/core/datasets.py:209
↓ 1 callersFunctionPoseRefine
(predictions, flowModelFile, class_number, meshes, mesh_diameters, render_K)
refinement.py:93
↓ 1 callersFunctionRefinePoseFromFlows
(exemplar_K, target_K, trans_M, patch_vertmaps, patch_forward_flows)
refinement.py:33
↓ 1 callersMethod_print_training_status
(self)
flow/train.py:107
↓ 1 callersMethodbasic_encoder
(self, x)
flow/core/extractor.py:250
↓ 1 callersFunctionbilinear_sampler
Wrapper for grid_sample, uses pixel coordinates
flow/core/utils/utils.py:57
↓ 1 callersFunctioncompute_pose_diff
(mesh3ds, K, gtR, gtT, predR, predT, isSym=False)
pose_utils.py:3
↓ 1 callersMethodcorr
(fmap1, fmap2)
flow/core/corr.py:53
↓ 1 callersFunctioncount_parameters
(model)
flow/train.py:87
↓ 1 callersFunctionevalute_auc_metric
(error_3ds, max_err)
pose_utils.py:35
↓ 1 callersFunctionfetch_dataloader
Create the data loader for the corresponding trainign set
flow/core/datasets.py:101
↓ 1 callersFunctionfetch_optimizer
Create the optimizer and learning rate scheduler
flow/train.py:91
↓ 1 callersFunctionflow_uv_to_colors
Applies the flow color wheel to (possibly clipped) flow components u and v. According to the C++ source code of Daniel Scharstein Accord
flow/core/utils/flow_viz.py:71
↓ 1 callersMethodforward
Estimate optical flow between pair of frames
flow/core/raft.py:89
↓ 1 callersFunctionget_available_memory
()
flow/core/utils/frame_utils.py:122
↓ 1 callersFunctionget_crop_M
(mask_xs, mask_ys, patch_width, patch_height)
flow/core/datasets.py:254
↓ 1 callersFunctionget_single_bop_annotation
(img_path, objID_2_clsID, mem_cache=None)
flow/core/datasets.py:143
↓ 1 callersMethodinitialize_flow
Flow is represented as difference between two coordinate grids flow = coords1 - coords0
flow/core/raft.py:66
↓ 1 callersFunctionload_bop_meshes
(model_path, device='cuda')
flow/core/datasets.py:29
↓ 1 callersFunctionload_image_cached
(img_path, mem_cache=None)
flow/core/datasets.py:123
↓ 1 callersFunctionload_image_cached
(img_path, mem_cache=None)
flow/core/utils/frame_utils.py:126
↓ 1 callersFunctionmake_colorwheel
Generates a color wheel for optical flow visualization as presented in: Baker et al. "A Database and Evaluation Methodology for Optical F
flow/core/utils/flow_viz.py:21
↓ 1 callersMethodpad
(self, *inputs)
flow/core/utils/utils.py:18
↓ 1 callersFunctionreadFlow
Read .flo file in Middlebury format
flow/core/utils/frame_utils.py:12
↓ 1 callersFunctionreadPFM
(file)
flow/core/utils/frame_utils.py:33
↓ 1 callersFunctionrender_objects_pytorch3d
(meshes, poses, K, w, h)
flow/core/datasets.py:54
↓ 1 callersMethodscale
(self, loss)
flow/train.py:44
↓ 1 callersFunctionsequence_loss
Loss function defined over sequence of flow predictions
flow/train.py:60
↓ 1 callersFunctiontrain
(args)
flow/train.py:146
↓ 1 callersMethodunscale_
(self, optimizer)
flow/train.py:46
↓ 1 callersFunctionupflow8
(flow, mode='bilinear')
flow/core/utils/utils.py:80
↓ 1 callersMethodupsample_flow
Upsample flow field [H/8, W/8, 2] -> [H, W, 2] using convex combination
flow/core/raft.py:75
↓ 1 callersMethodwrite_dict
(self, results)
flow/train.py:135
Method__call__
(self, coords)
flow/core/corr.py:29
Method__call__
(self, coords)
flow/core/corr.py:74
Method__enter__
(self)
flow/core/raft.py:18
Method__exit__
(self, *args)
flow/core/raft.py:20
Method__getitem__
(self, index)
flow/core/datasets.py:324
Method__init__
(self, enabled=None)
flow/train.py:42
Method__init__
(self, model, scheduler)
flow/train.py:100
Method__init__
(self, start, end, n_iter, anneal_fn, params = None)
flow/scheduler.py:38
Method__init__
( self, optimizer, lr_base, max_iter, strategy, params,
flow/scheduler.py:58
Method__init__
(self, in_planes, planes, norm_fn='group', stride=1)
flow/core/extractor.py:8
Method__init__
(self, in_planes, planes, norm_fn='group', stride=1)
flow/core/extractor.py:62
Method__init__
(self, input_dim=3, output_dim=128, waveLevelCnt=5, norm_fn='batch', dropout=0.0)
flow/core/extractor.py:196
Method__init__
(self, output_dim=128, norm_fn='batch', dropout=0.0)
flow/core/extractor.py:297
Method__init__
(self, render_K, image_list_file='', mesh_path='')
flow/core/datasets.py:305
Method__init__
(self, fmap1, fmap2, num_levels=4, radius=4)
flow/core/corr.py:13
Method__init__
(self, fmap1, fmap2, num_levels=4, radius=4)
flow/core/corr.py:64
Method__init__
(self, input_dim=128, hidden_dim=256)
flow/core/update.py:6
Method__init__
(self, hidden_dim=128, input_dim=192+128)
flow/core/update.py:33
Method__init__
(self, args)
flow/core/update.py:62
Method__init__
(self, args)
flow/core/update.py:79
Method__init__
(self, args, hidden_dim=96)
flow/core/update.py:99
Method__init__
(self, args, hidden_dim=128, input_dim=128)
flow/core/update.py:114
Method__init__
(self, enabled)
flow/core/raft.py:16
Method__init__
(self, args)
flow/core/raft.py:24
Method__init__
(self, dims, mode='sintel')
flow/core/utils/utils.py:9
Method__len__
(self)
flow/core/datasets.py:371
Method__rmul__
(self, v)
flow/core/datasets.py:366
Functionanneal_cos
(start, end, proportion, params=None)
flow/scheduler.py:17
Functionanneal_cospow
(start, end, proportion, params=None)
flow/scheduler.py:22
Functionanneal_linear
(start, end, proportion, params=None)
flow/scheduler.py:5
Functionanneal_multi_steps
(start, end, proportion, params)
flow/scheduler.py:8
Functionanneal_poly
(start, end, proportion, power=0.9)
flow/scheduler.py:31
Functionanneal_tanh
(start, end, proportion, lower=-6, upper=3)
flow/scheduler.py:34
Functiondraw_bounding_box
(cvImg, R, T, mesh, intrinsics, color)
flow/core/utils/pose_viz.py:5
Functiondraw_pose_axis
(cvImg, R, T, mesh, intrinsics)
flow/core/utils/pose_viz.py:18
Functiondraw_pose_contour
(cvImg, R, T, mesh, K, color)
flow/core/utils/pose_viz.py:42
Functiondraw_z_axis
(cvImg, R, T, mesh, intrinsics, color)
flow/core/utils/pose_viz.py:31
Functionflow_to_image
Expects a two dimensional flow image of shape. Args: flow_uv (np.ndarray): Flow UV image of shape [H,W,2] clip_flow (float,
flow/core/utils/flow_viz.py:110
Methodforward
(self, x)
flow/core/extractor.py:49
Methodforward
(self, x)
flow/core/extractor.py:108
Methodforward
(self, x)
flow/core/extractor.py:169
Methodforward
(self, x)
flow/core/extractor.py:275
Methodforward
(self, x)
flow/core/extractor.py:345
Methodforward
(self, x)
flow/core/update.py:12
Methodforward
(self, h, x)
flow/core/update.py:22
Methodforward
(self, h, x)
flow/core/update.py:44
Methodforward
(self, flow, corr)
flow/core/update.py:70
Methodforward
(self, flow, corr)
flow/core/update.py:88
Methodforward
(self, net, inp, corr, flow)
flow/core/update.py:105
Methodforward
(self, net, inp, corr, flow, upsample=True)
flow/core/update.py:126
next →1–100 of 115, ranked by callers