ppyolo_r50vd_dcn_antiuav.yml

architecture: YOLOv3
pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/ResNet50_vd_ssld_pretrained.pdparams
norm_type: sync_bn
use_ema: true
ema_decay: 0.9998

YOLOv3:
  backbone: ResNet
  neck: PPYOLOFPN
  yolo_head: YOLOv3Head
  post_process: BBoxPostProcess

ResNet:
  depth: 50
  variant: d
  return_idx: [1, 2, 3]
  dcn_v2_stages: [3]
  freeze_at: -1
  freeze_norm: false
  norm_decay: 0.

PPYOLOFPN:
  coord_conv: true
  drop_block: true
  block_size: 3
  keep_prob: 0.9
  spp: true

YOLOv3Head:
  anchors: [[6, 6], [10, 10], [15, 16],
            [23, 13], [21, 19], [32, 16],
            [31, 27], [44, 20], [52, 36]]
  anchor_masks: [[6, 7, 8], [3, 4, 5], [0, 1, 2]]
  loss: YOLOv3Loss
  iou_aware: true
  iou_aware_factor: 0.4

YOLOv3Loss:
  ignore_thresh: 0.7
  downsample: [32, 16, 8]
  label_smooth: false
  scale_x_y: 1.05
  iou_loss: IouLoss
  iou_aware_loss: IouAwareLoss

IouLoss:
  loss_weight: 2.5
  loss_square: true

IouAwareLoss:
  loss_weight: 1.0

BBoxPostProcess:
  decode:
    name: YOLOBox
    conf_thresh: 0.01
    downsample_ratio: 32
    clip_bbox: true
    scale_x_y: 1.05
  nms:
    name: MatrixNMS
    keep_top_k: 100
    score_threshold: 0.01
    post_threshold: 0.01
    nms_top_k: -1
    background_label: -1