v3 (iou loss, Normalizer: (iou: 0.07, obj: 1.00, cls: 1.00) Region 139 Avg (IOU: 0.754252), count: 12, class_loss = 11.161986, iou_loss = 276.395813, total_loss = 287.557800
v3 (iou loss, Normalizer: (iou: 0.07, obj: 1.00, cls: 1.00) Region 150 Avg (IOU: 0.772841), count: 9, class_loss = 6.131786, iou_loss = 52.778233, total_loss = 58.910019
v3 (iou loss, Normalizer: (iou: 0.07, obj: 1.00, cls: 1.00) Region 161 Avg (IOU: 0.000000), count: 1, class_loss = 0.509116, iou_loss = 0.000000, total_loss = 0.509116 total_bbox = 109507, rewritten_bbox = 0.012785 %
v3 (iou loss, Normalizer: (iou: 0.07, obj: 1.00, cls: 1.00) Region 139 Avg (IOU: 0.771520), count: 8, class_loss = 4.944231, iou_loss = 214.195007, total_loss = 219.139236
v3 (iou loss, Normalizer: (iou: 0.07, obj: 1.00, cls: 1.00) Region 150 Avg (IOU: 0.689796), count: 6, class_loss = 6.309541, iou_loss = 22.156059, total_loss = 28.465599
v3 (iou loss, Normalizer: (iou: 0.07, obj: 1.00, cls: 1.00) Region 161 Avg (IOU: 0.842883), count: 1, class_loss = 0.827105, iou_loss = 5.044473, total_loss = 5.871578 total_bbox = 109522, rewritten_bbox = 0.012783 %
Avg IOU(IOU) : 当前迭代中,预测的box与标注的box的平均交并比,越大越好,期望数值为1.
count:当前送入训练的图片中,正样本的图片的数量。
iou_loss:iou的损失值,越小越好
total_loss:总损失值,越小越好