FocalLoss Save

Caffe implementation of FAIR paper "Focal Loss for Dense Object Detection" for SSD.

Project README

FocalLoss

Caffe implementation of FAIR paper "Focal Loss for Dense Object Detection" for SSD.

layer {
  name: "mbox_loss"
  type: "MultiBoxFocalLoss" #change the type
  bottom: "mbox_loc"
  bottom: "mbox_conf"
  bottom: "mbox_priorbox"
  bottom: "label"
  top: "mbox_loss"
  include {
    phase: TRAIN
  }
  propagate_down: true
  propagate_down: true
  propagate_down: false
  propagate_down: false
  loss_param {
    normalization: VALID
  }
  focal_loss_param { #set the alpha and gamma, default is alpha=0.25, gamma=2.0
    alpha: 0.25
    gamma: 2.0
  }
  multibox_loss_param {
    loc_loss_type: SMOOTH_L1
    conf_loss_type: SOFTMAX
    loc_weight: 1.0
    num_classes: 21
    share_location: true
    match_type: PER_PREDICTION
    overlap_threshold: 0.5
    use_prior_for_matching: true
    background_label_id: 0
    use_difficult_gt: true
    neg_pos_ratio: 3.0
    neg_overlap: 0.5
    code_type: CENTER_SIZE
    ignore_cross_boundary_bbox: false
    mining_type: NONE #do not use OHEM
  }
}
Open Source Agenda is not affiliated with "FocalLoss" Project. README Source: chuanqi305/FocalLoss
Stars
172
Open Issues
9
Last Commit
6 years ago
License
MIT
Tags

Open Source Agenda Badge

Open Source Agenda Rating