Python mmdet.core.build_assigner() Examples

The following are 3 code examples of mmdet.core.build_assigner(). You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may also want to check out all available functions/classes of the module mmdet.core , or try the search function .
Example #1
Source File: standard_roi_head.py    From mmdetection with Apache License 2.0 5 votes vote down vote up 
def init_assigner_sampler(self):
        """Initialize assigner and sampler."""
        self.bbox_assigner = None
        self.bbox_sampler = None
        if self.train_cfg:
            self.bbox_assigner = build_assigner(self.train_cfg.assigner)
            self.bbox_sampler = build_sampler(
                self.train_cfg.sampler, context=self)
Example #2
Source File: test_heads.py    From mmdetection with Apache License 2.0 5 votes vote down vote up 
def _dummy_bbox_sampling(proposal_list, gt_bboxes, gt_labels):
    """Create sample results that can be passed to BBoxHead.get_targets."""
    num_imgs = 1
    feat = torch.rand(1, 1, 3, 3)
    assign_config = dict(
        type='MaxIoUAssigner',
        pos_iou_thr=0.5,
        neg_iou_thr=0.5,
        min_pos_iou=0.5,
        ignore_iof_thr=-1)
    sampler_config = dict(
        type='RandomSampler',
        num=512,
        pos_fraction=0.25,
        neg_pos_ub=-1,
        add_gt_as_proposals=True)
    bbox_assigner = build_assigner(assign_config)
    bbox_sampler = build_sampler(sampler_config)
    gt_bboxes_ignore = [None for _ in range(num_imgs)]
    sampling_results = []
    for i in range(num_imgs):
        assign_result = bbox_assigner.assign(proposal_list[i], gt_bboxes[i],
                                             gt_bboxes_ignore[i], gt_labels[i])
        sampling_result = bbox_sampler.sample(
            assign_result,
            proposal_list[i],
            gt_bboxes[i],
            gt_labels[i],
            feats=feat)
        sampling_results.append(sampling_result)

    return sampling_results
Example #3
Source File: single_stage.py    From AugFPN with Apache License 2.0 4 votes vote down vote up 
def forward_train(self,
                      img,
                      img_metas,
                      gt_bboxes,
                      gt_labels,
                      gt_bboxes_ignore=None):
        if self.use_consistent_supervision:
            x, y = self.extract_feat(img)
            gt_bboxes_auxiliary = [gt.clone() for gt in gt_bboxes]
            gt_labels_auxiliary = [label.clone() for label in gt_labels]
        else:
            x = self.extract_feat(img)
        outs = self.bbox_head(x)
        loss_inputs = outs + (gt_bboxes, gt_labels, img_metas, self.train_cfg)

        losses = self.bbox_head.loss(
            *loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore)
        if self.use_consistent_supervision:
            proposal_cfg = self.train_cfg.auxiliary.proposal
            proposal_inputs = outs + (img_metas, proposal_cfg)
            proposal_list = self.bbox_head.get_bboxes_auxiliary(*proposal_inputs)
        
            bbox_assigner = build_assigner(self.train_cfg.auxiliary.assigner)
            bbox_sampler = build_sampler(
                self.train_cfg.auxiliary.sampler, context=self)
            num_imgs = img.size(0)
            if gt_bboxes_ignore is None:
                gt_bboxes_ignore = [None for _ in range(num_imgs)]
            sampling_results = []
            for i in range(num_imgs):

                assign_result = bbox_assigner.assign(
                    proposal_list[i], gt_bboxes_auxiliary[i], gt_bboxes_ignore[i],
                    gt_labels_auxiliary[i])
                sampling_result = bbox_sampler.sample(
                    assign_result,
                    proposal_list[i],
                    gt_bboxes_auxiliary[i],
                    gt_labels_auxiliary[i],
                    feats=[lvl_feat[i][None] for lvl_feat in x])
                sampling_results.append(sampling_result)
            rois = bbox2roi([res.bboxes for res in sampling_results])
            bbox_feats_raw = self.auxiliary_bbox_roi_extractor(y[:self.auxiliary_bbox_roi_extractor.num_inputs], rois)
            cls_score_auxiliary, bbox_pred_auxiliary = self.auxiliary_bbox_head(bbox_feats_raw)

            bbox_targets = self.auxiliary_bbox_head.get_target(
                sampling_results, gt_bboxes, gt_labels, self.train_cfg.auxiliary.rcnn)
        
            loss_bbox_auxiliary = self.auxiliary_bbox_head.loss(cls_score_auxiliary, bbox_pred_auxiliary,
                                            *bbox_targets, alpha=0.25, num_level=3)   
            losses.update(loss_bbox_auxiliary) 

        return losses