MCPcopy Create free account

hub / github.com/ZwwWayne/K-Net / functions

Functions71 in github.com/ZwwWayne/K-Net

⨍ Functions 71◇ Types & classes 17
↓ 7 callersMethodinfo
Returns a dictionary of info about the object.
knet/det/mask_pseudo_sampler.py:63
↓ 3 callersMethod__init__
(self, **kwargs)
knet/seg/seg_kernel_head.py:115
↓ 3 callersMethod_decode_init_proposals
(self, img, img_metas)
knet/det/kernel_head.py:200
↓ 3 callersMethod_mask_forward
(self, stage, x, object_feats, mask_preds, img_metas)
knet/det/kernel_iter_head.py:118
↓ 2 callersMethodassign
Computes one-to-one matching based on the weighted costs. This method assign each query prediction to a ground truth or background. T
knet/det/mask_hungarian_assigner.py:157
↓ 2 callersMethodget_targets
(self, sampling_results, gt_mask, rpn_train_cfg,
knet/det/kernel_head.py:466
↓ 2 callersMethodloss
(self, mask_pred, cls_scores, seg_preds, proposal_feats,
knet/det/kernel_head.py:336
↓ 2 callersMethodrescale_masks
(self, masks_per_img, img_meta)
knet/det/kernel_update_head.py:442
↓ 2 callersMethodsample
Directly returns the positive and negative indices of samples. Args: assign_result (:obj:`AssignResult`): Assigned results
knet/det/mask_pseudo_sampler.py:91
↓ 2 callersMethodsimple_test_rpn
Forward function in testing stage.
knet/det/kernel_head.py:503
↓ 1 callersMethod_init_layers
Initialize a sparse set of proposal boxes and proposal features.
knet/det/kernel_head.py:122
↓ 1 callersMethoddice_loss
(cls, input, target, eps=1e-3)
knet/det/mask_hungarian_assigner.py:44
↓ 1 callersMethodgenerate_coord
(self, input_feat)
knet/det/semantic_fpn_wrapper.py:179
↓ 1 callersMethodget_panoptic
(self, cls_scores, mask_preds, test_cfg, img_meta)
knet/det/kernel_iter_head.py:303
↓ 1 callersMethodget_seg_masks
(self, masks_per_img, labels_per_img, scores_per_img, test_cfg, img_meta)
knet/det/kernel_update_head.py:459
↓ 1 callersMethodmerge_stuff_thing
(self, total_masks, total_labels,
knet/det/kernel_iter_head.py:322
↓ 1 callersMethodsegm2result
(self, mask_preds, det_labels, cls_scores)
knet/det/kernel_update_head.py:468
↓ 1 callersFunctionsem2ins_masks
Convert semantic segmentation mask to binary masks Args: gt_sem_seg (torch.Tensor): Semantic masks to be converted. [0, num_t
knet/det/utils.py:4
Method__call__
Args: bbox_pred (Tensor): Predicted boxes with normalized coordinates (cx, cy, w, h), which are all in range [0,
knet/det/mask_hungarian_assigner.py:56
Method__call__
Args: cls_pred (Tensor): Predicted classification logits, shape [num_query, num_class]. gt_labels (Te
knet/det/mask_hungarian_assigner.py:89
Method__init__
(self, in_channels=256, feat_channels=64, out_channels=None
knet/kernel_updator.py:10
Method__init__
(self, **kwargs)
knet/seg/seg_kernel_head.py:19
Method__init__
(self, **kwargs)
knet/seg/seg_kernel_head.py:54
Method__init__
(self, **kwargs)
knet/seg/seg_kernel_head.py:136
Method__init__
(self, num_stages, kernel_generate_head, kernel_update_head
knet/seg/iter_decode_head.py:18
Method__init__
(self, num_classes=80, num_ffn_fcs=2, num_heads=8,
knet/seg/kernel_update_head.py:14
Method__init__
(self, weight=1., pred_act=False, act_mode='sigmoid',
knet/det/mask_hungarian_assigner.py:34
Method__init__
(self, weight=1., pred_act=False, act_mode='sigmoid')
knet/det/mask_hungarian_assigner.py:84
Method__init__
(self, cls_cost=dict(type='ClassificationCost', weight=1.), mask_cost=dict(t
knet/det/mask_hungarian_assigner.py:142
Method__init__
(self, num_proposals=100, in_channels=256, out_channels=256
knet/det/kernel_head.py:14
Method__init__
(self, num_classes=80, num_ffn_fcs=2, num_heads=8,
knet/det/kernel_update_head.py:21
Method__init__
(self, *args, num_thing_classes=80, num_stuff_classes=53,
knet/det/knet.py:13
Method__init__
(self, num_stages=6, recursive=False, assign_stages=5,
knet/det/kernel_iter_head.py:15
Method__init__
(self, in_channels, feat_channels, out_channels,
knet/det/semantic_fpn_wrapper.py:25
Method__init__
(self, pos_inds, neg_inds, masks, gt_masks, assign_result, gt_flags)
knet/det/mask_pseudo_sampler.py:26
Method__init__
(self, **kwargs)
knet/det/mask_pseudo_sampler.py:80
Method__nice__
(self)
knet/det/mask_pseudo_sampler.py:54
Method_get_target_single
(self, pos_inds, neg_inds, pos_mask, neg_mask, pos_gt_mask, pos_gt_labels, gt_sem_s
knet/det/kernel_head.py:427
Method_get_target_single
(self, pos_inds, neg_inds, pos_mask, neg_mask, pos_gt_mask, pos_gt_labels, gt_sem_s
knet/det/kernel_update_head.py:350
Method_sample_neg
Sample negative samples.
knet/det/mask_pseudo_sampler.py:87
Method_sample_pos
Sample positive samples.
knet/det/mask_pseudo_sampler.py:83
Methodaug_test
(self, features, proposal_list, img_metas, rescale=False)
knet/det/kernel_iter_head.py:285
Methodforward
(self, update_feature, input_feature)
knet/kernel_updator.py:56
Methodforward
Forward function.
knet/seg/seg_kernel_head.py:22
Methodforward
Forward function.
knet/seg/seg_kernel_head.py:57
Methodforward
Forward function.
knet/seg/seg_kernel_head.py:118
Methodforward
Forward function.
knet/seg/seg_kernel_head.py:139
Methodforward
Forward function.
knet/seg/iter_decode_head.py:38
Methodforward
Forward function of Dynamic Instance Interactive Head. Args: x (Tensor): Feature map from FPN with shape (batch_s
knet/seg/kernel_update_head.py:110
Methodforward
(self, x, proposal_feat, mask_preds, prev_cls_
knet/det/kernel_update_head.py:172
Methodforward
(self, inputs)
knet/det/semantic_fpn_wrapper.py:190
Methodforward_dummy
Dummy forward function. Used in flops calculation.
knet/det/kernel_head.py:507
Methodforward_dummy
Used for computing network flops. See `mmdetection/tools/get_flops.py`
knet/det/knet.py:133
Methodforward_dummy
Dummy forward function when do the flops computing.
knet/det/kernel_iter_head.py:288
Methodforward_train
Forward function in training stage.
knet/det/kernel_head.py:266
Methodforward_train
(self, img, img_metas, gt_bboxes=None,
knet/det/knet.py:29
Methodforward_train
(self, x, proposal_feats, mask_preds,
knet/det/kernel_iter_head.py:139
Methodget_targets
(self, sampling_results, gt_mask, gt_labels,
knet/det/kernel_update_head.py:405
Methodinit_assigner_sampler
Initialize assigner and sampler for each stage.
knet/det/kernel_iter_head.py:85
Methodinit_bbox_head
Initialize box head and box roi extractor. Args: mask_roi_extractor (dict): Config of box roi extractor. mask_head (d
knet/det/kernel_iter_head.py:76
Methodinit_mask_head
Initialize mask head and mask roi extractor. Args: mask_roi_extractor (dict): Config of mask roi extractor. mask_head
knet/det/kernel_iter_head.py:101
Methodinit_weights
Use xavier initialization for all weight parameter and set classification head bias as a specific value when use focal loss.
knet/seg/kernel_update_head.py:94
Methodinit_weights
(self)
knet/det/kernel_head.py:173
Methodinit_weights
Use xavier initialization for all weight parameter and set classification head bias as a specific value when use focal loss.
knet/det/kernel_update_head.py:153
Methodinit_weights
(self)
knet/det/kernel_iter_head.py:97
Methodinit_weights
(self)
knet/det/semantic_fpn_wrapper.py:172
Methodloss
(self, object_feats, cls_score, mask_pred, labels,
knet/det/kernel_update_head.py:279
Methodlosses
(self, seg_logit, seg_label)
knet/seg/iter_decode_head.py:52
Methodmasks
torch.Tensor: concatenated positive and negative boxes
knet/det/mask_pseudo_sampler.py:50
Methodsimple_test
(self, img, img_metas, rescale=False)
knet/det/knet.py:118
Methodsimple_test
(self, x, proposal_feats, mask_preds,
knet/det/kernel_iter_head.py:233