↓ 5 callersFunctioncheck_points_inside_bboxesr""" Args: points (Tensor, float32): shape[L, 2], "xy" format, L: num_anchors bboxes (Tensor, float32): shape[B, n, 4], "xmin, ymi
ppdet/modeling/assigners/utils.py:108
↓ 4 callersMethod__init__(self, channel, lr_mult, conv_decay, reduction=4, name="")
ppdet/modeling/backbones/mobilenet_v3.py:185
↓ 4 callersMethod__init__(
self,
arch='focalnet_T_224_1k_srf',
out_indices=(0, 1, 2, 3),
ppdet/modeling/backbones/focalnet.py:571
↓ 4 callersFunctioncocoapi_eval Args: jsonfile (str): Evaluation json file, eg: bbox.json, mask.json. style (str): COCOeval style, can be `bbox` , `segm` , `prop
ppdet/metrics/coco_utils.py:70
↓ 4 callersFunctioncompute_max_iou_anchorr""" For each anchor, find the GT with the largest IOU. Args: ious (Tensor, float32): shape[B, n, L], n: num_gts, L: num_anchors R
ppdet/modeling/assigners/utils.py:153
↓ 4 callersMethodforwardr""" Args: boxes (Tensor): [l, b, query, 4] logits (Tensor): [l, b, query, num_classes] gt_bbox (List(Tens
ppdet/modeling/losses/detr_loss.py:319
↓ 4 callersFunctionnms Apply NMS to avoid detecting too many overlapping bounding boxes. Args: dets: shape [N, 5], [score, x1, y1, x2, y2] m
deploy/python/utils.py:305
↓ 4 callersFunctionxywh2xyxy [x, y, w, h] to [x1, y1, x2, y2], paddle Tensor op
ppdet/modeling/losses/yolov7_loss.py:673
↓ 3 callersMethod_bbox_loss(self, pred_dist, pred_bboxes, anchor_points, assigned_labels,
assigned_bboxes, assigned_sc
ppdet/modeling/heads/yolov6_head.py:289