configs/cse/Base-DensePose-RCNN-FPN.yaml

VERSION: 2
MODEL:
  META_ARCHITECTURE: "GeneralizedRCNN"
  BACKBONE:
    NAME: "build_resnet_fpn_backbone"
  RESNETS:
    OUT_FEATURES: ["res2", "res3", "res4", "res5"]
  FPN:
    IN_FEATURES: ["res2", "res3", "res4", "res5"]
  ANCHOR_GENERATOR:
    SIZES: [[32], [64], [128], [256], [512]]  # One size for each in feature map
    ASPECT_RATIOS: [[0.5, 1.0, 2.0]]  # Three aspect ratios (same for all in feature maps)
  RPN:
    IN_FEATURES: ["p2", "p3", "p4", "p5", "p6"]
    PRE_NMS_TOPK_TRAIN: 2000  # Per FPN level
    PRE_NMS_TOPK_TEST: 1000  # Per FPN level
    # Detectron1 uses 2000 proposals per-batch,
    # (See "modeling/rpn/rpn_outputs.py" for details of this legacy issue)
    # which is approximately 1000 proposals per-image since the default batch size for FPN is 2.
    POST_NMS_TOPK_TRAIN: 1000
    POST_NMS_TOPK_TEST: 1000

  DENSEPOSE_ON: True
  ROI_HEADS:
    NAME: "DensePoseROIHeads"
    IN_FEATURES: ["p2", "p3", "p4", "p5"]
    NUM_CLASSES: 1
  ROI_BOX_HEAD:
    NAME: "FastRCNNConvFCHead"
    NUM_FC: 2
    POOLER_RESOLUTION: 7
    POOLER_SAMPLING_RATIO: 2
    POOLER_TYPE: "ROIAlign"
  ROI_DENSEPOSE_HEAD:
    NAME: "DensePoseV1ConvXHead"
    POOLER_TYPE: "ROIAlign"
    NUM_COARSE_SEGM_CHANNELS: 2
    PREDICTOR_NAME: "DensePoseEmbeddingPredictor"
    LOSS_NAME: "DensePoseCseLoss"
    CSE:
      # embedding loss, possible values:
      # - "EmbeddingLoss"
      # - "SoftEmbeddingLoss"
      EMBED_LOSS_NAME: "EmbeddingLoss"
SOLVER:
  IMS_PER_BATCH: 16
  BASE_LR: 0.01
  STEPS: (60000, 80000)
  MAX_ITER: 90000
  WARMUP_FACTOR: 0.1
  CLIP_GRADIENTS:
    CLIP_TYPE: norm
    CLIP_VALUE: 1.0
    ENABLED: true
    NORM_TYPE: 2.0
INPUT:
  MIN_SIZE_TRAIN: (640, 672, 704, 736, 768, 800)
DENSEPOSE_EVALUATION:
  TYPE: cse
  STORAGE: file