使用CIOU作为YOLOv3的定位损失，focal loss作为分类损失

def box_ciou(b1, b2):
    """
    输入为：    ----------
    b1: tensor, shape=(batch, feat_w, feat_h, anchor_num, 4), xywh
    b2: tensor, shape=(batch, feat_w, feat_h, anchor_num, 4), xywh
    返回为：    -------
    ciou: tensor, shape=(batch, feat_w, feat_h, anchor_num, 1)
    """
    # 求出预测框左上角右下角
    b1_xy = b1[..., :2]
    b1_wh = b1[..., 2:4]
    b1_wh_half = b1_wh/2.
    b1_mins = b1_xy - b1_wh_half
    b1_maxes = b1_xy + b1_wh_half
    # 求出真实框左上角右下角
    b2_xy = b2[..., :2]
    b2_wh = b2[..., 2:4]
    b2_wh_half = b2_wh/2.
    b2_mins = b2_xy - b2_wh_half
    b2_maxes = b2_xy + b2_wh_half
 
    # 求真实框和预测框所有的iou
    intersect_mins = torch.max(b1_mins, b2_mins)
    intersect_maxes = torch.min(b1_maxes, b2_maxes)
    intersect_wh = torch.max(intersect_maxes - intersect_mins, torch.zeros_like(intersect_maxes))
    intersect_area = intersect_wh[..., 0] * intersect_wh[..., 1]
    b1_area = b1_wh[..., 0] * b1_wh[..., 1]
    b2_area = b2_wh[..., 0] * b2_wh[..., 1]
    union_area = b1_area + b2_area - intersect_area
    iou = intersect_area / (union_area + 1e-6)
 
    # 计算中心的差距
    center_distance = torch.sum(torch.pow((b1_xy - b2_xy), 2), axis=-1)
    # 找到包裹两个框的最小框的左上角和右下角
    enclose_mins = torch.min(b1_mins, b2_mins)
    enclose_maxes = torch.max(b1_maxes, b2_maxes)
    enclose_wh = torch.max(enclose_maxes - enclose_mins, torch.zeros_like(intersect_maxes))
    # 计算对角线距离
    enclose_diagonal = torch.sum(torch.pow(enclose_wh,2), axis=-1)
    ciou = iou - 1.0 * (center_distance) / (enclose_diagonal + 1e-7)
    v = (4 / (math.pi ** 2)) * torch.pow((torch.atan(b1_wh[..., 0]/b1_wh[..., 1]) - torch.atan(b2_wh[..., 0]/b2_wh[..., 1])), 2)
    alpha = v / (1.0 - iou + v)
    ciou = ciou - alpha * v
    return ciou



box_loss_scale = 2 - y_true[l][...,2:3]*y_true[l][...,3:4]
# raw_true_xy = [13x(x_true-1),13x(y_true-1)] (< 0)
# raw_pred[..., 0:2] = [pred_x,pred_y]
# binary_crossentropy = [13x(x_true-1),13x(y_true-1)]xlog([pred_x,pred_y])+(1-[13x(x_true-1),13x(y_true-1)])xlog(1-[pred_x,pred_y])
xy_loss = object_mask * box_loss_scale 
		* K.binary_crossentropy(raw_true_xy, raw_pred[...,0:2], from_logits=True)
wh_loss = object_mask * box_loss_scale *
	 0.5 * K.square(raw_true_wh-raw_pred[...,2:4])
confidence_loss = object_mask * K.binary_crossentropy(object_mask, raw_pred[...,4:5], from_logits=True)+ 
			(1-object_mask) * K.binary_crossentropy(object_mask, raw_pred[...,4:5], from_logits=True) * ignore_mask
class_loss = object_mask * K.binary_crossentropy(true_class_probs, raw_pred[...,5:], from_logits=True)
xy_loss = K.sum(xy_loss) / mf
wh_loss = K.sum(wh_loss) / mf
confidence_loss = K.sum(confidence_loss) / mf
class_loss = K.sum(class_loss) / mf
loss += xy_loss + wh_loss + confidence_loss + class_loss


box_loss_scale = 2 - y_true[l][...,2:3]*y_true[l][...,3:4]
raw_true_box = y_true[l][...,0:4]
ciou = box_ciou(pred_box, raw_true_box)
ciou_loss = object_mask * box_loss_scale * (1 - ciou)
ciou_loss = K.sum(ciou_loss) / mf
location_loss = ciou_loss
confidence_loss = object_mask * K.binary_crossentropy(object_mask, raw_pred[...,4:5], from_logits=True)+(1-object_mask) * K.binary_crossentropy(object_mask, raw_pred[...,4:5], from_logits=True) * ignore_mask
class_loss = object_mask * K.binary_crossentropy(true_class_probs, raw_pred[...,5:], from_logits=True)
confidence_loss = K.sum(confidence_loss) / mf
class_loss = K.sum(class_loss) / mf
loss += location_loss + confidence_loss + class_loss

更多内容请关注公众号，了解嵌入式ai前沿资讯！