Loss and IoU speed improvements (#7361)

utils/loss.py

@@ -138,7 +138,7 @@ class ComputeLoss:
                 pxy = pxy.sigmoid() * 2 - 0.5
                 pwh = (pwh.sigmoid() * 2) ** 2 * anchors[i]
                 pbox = torch.cat((pxy, pwh), 1)  # predicted box
-                iou = bbox_iou(pbox.T, tbox[i], x1y1x2y2=False, CIoU=True)  # iou(prediction, target)
+                iou = bbox_iou(pbox, tbox[i], CIoU=True).squeeze()  # iou(prediction, target)
                 lbox += (1.0 - iou).mean()  # iou loss
 
                 # Objectness
@@ -180,7 +180,7 @@ class ComputeLoss:
         tcls, tbox, indices, anch = [], [], [], []
         gain = torch.ones(7, device=self.device)  # normalized to gridspace gain
         ai = torch.arange(na, device=self.device).float().view(na, 1).repeat(1, nt)  # same as .repeat_interleave(nt)
-        targets = torch.cat((targets.repeat(na, 1, 1), ai[:, :, None]), 2)  # append anchor indices
+        targets = torch.cat((targets.repeat(na, 1, 1), ai[..., None]), 2)  # append anchor indices
 
         g = 0.5  # bias
         off = torch.tensor(
@@ -199,10 +199,10 @@ class ComputeLoss:
             gain[2:6] = torch.tensor(p[i].shape)[[3, 2, 3, 2]]  # xyxy gain
 
             # Match targets to anchors
-            t = targets * gain
+            t = targets * gain  # shape(3,n,7)
             if nt:
                 # Matches
-                r = t[:, :, 4:6] / anchors[:, None]  # wh ratio
+                r = t[..., 4:6] / anchors[:, None]  # wh ratio
                 j = torch.max(r, 1 / r).max(2)[0] < self.hyp['anchor_t']  # compare
                 # j = wh_iou(anchors, t[:, 4:6]) > model.hyp['iou_t']  # iou(3,n)=wh_iou(anchors(3,2), gwh(n,2))
                 t = t[j]  # filter

utils/metrics.py

@@ -206,37 +206,36 @@ class ConfusionMatrix:
             print(' '.join(map(str, self.matrix[i])))
 
 
-def bbox_iou(box1, box2, x1y1x2y2=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-7):
-    # Returns the IoU of box1 to box2. box1 is 4, box2 is nx4
-    box2 = box2.T
+def bbox_iou(box1, box2, xywh=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-7):
+    # Returns Intersection over Union (IoU) of box1(1,4) to box2(n,4)
 
     # Get the coordinates of bounding boxes
-    if x1y1x2y2:  # x1, y1, x2, y2 = box1
-        b1_x1, b1_y1, b1_x2, b1_y2 = box1[0], box1[1], box1[2], box1[3]
-        b2_x1, b2_y1, b2_x2, b2_y2 = box2[0], box2[1], box2[2], box2[3]
-    else:  # transform from xywh to xyxy
-        b1_x1, b1_x2 = box1[0] - box1[2] / 2, box1[0] + box1[2] / 2
-        b1_y1, b1_y2 = box1[1] - box1[3] / 2, box1[1] + box1[3] / 2
-        b2_x1, b2_x2 = box2[0] - box2[2] / 2, box2[0] + box2[2] / 2
-        b2_y1, b2_y2 = box2[1] - box2[3] / 2, box2[1] + box2[3] / 2
+    if xywh:  # transform from xywh to xyxy
+        (x1, y1, w1, h1), (x2, y2, w2, h2) = box1.chunk(4, 1), box2.chunk(4, 1)
+        w1_, h1_, w2_, h2_ = w1 / 2, h1 / 2, w2 / 2, h2 / 2
+        b1_x1, b1_x2, b1_y1, b1_y2 = x1 - w1_, x1 + w1_, y1 - h1_, y1 + h1_
+        b2_x1, b2_x2, b2_y1, b2_y2 = x2 - w2_, x2 + w2_, y2 - h2_, y2 + h2_
+    else:  # x1, y1, x2, y2 = box1
+        b1_x1, b1_y1, b1_x2, b1_y2 = box1.chunk(4, 1)
+        b2_x1, b2_y1, b2_x2, b2_y2 = box2.chunk(4, 1)
+        w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps
+        w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps
 
     # Intersection area
     inter = (torch.min(b1_x2, b2_x2) - torch.max(b1_x1, b2_x1)).clamp(0) * \
             (torch.min(b1_y2, b2_y2) - torch.max(b1_y1, b2_y1)).clamp(0)
 
     # Union Area
-    w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps
-    w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps
     union = w1 * h1 + w2 * h2 - inter + eps
 
+    # IoU
     iou = inter / union
     if CIoU or DIoU or GIoU:
         cw = torch.max(b1_x2, b2_x2) - torch.min(b1_x1, b2_x1)  # convex (smallest enclosing box) width
         ch = torch.max(b1_y2, b2_y2) - torch.min(b1_y1, b2_y1)  # convex height
         if CIoU or DIoU:  # Distance or Complete IoU https://arxiv.org/abs/1911.08287v1
             c2 = cw ** 2 + ch ** 2 + eps  # convex diagonal squared
-            rho2 = ((b2_x1 + b2_x2 - b1_x1 - b1_x2) ** 2 +
-                    (b2_y1 + b2_y2 - b1_y1 - b1_y2) ** 2) / 4  # center distance squared
+            rho2 = ((b2_x1 + b2_x2 - b1_x1 - b1_x2) ** 2 + (b2_y1 + b2_y2 - b1_y1 - b1_y2) ** 2) / 4  # center dist ** 2
             if CIoU:  # https://github.com/Zzh-tju/DIoU-SSD-pytorch/blob/master/utils/box/box_utils.py#L47
                 v = (4 / math.pi ** 2) * torch.pow(torch.atan(w2 / h2) - torch.atan(w1 / h1), 2)
                 with torch.no_grad():
@@ -248,6 +247,11 @@ def bbox_iou(box1, box2, x1y1x2y2=True, GIoU=False, DIoU=False, CIoU=False, eps=
     return iou  # IoU
 
 
+def box_area(box):
+    # box = xyxy(4,n)
+    return (box[2] - box[0]) * (box[3] - box[1])
+
+
 def box_iou(box1, box2):
     # https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py
     """
@@ -261,16 +265,12 @@ def box_iou(box1, box2):
             IoU values for every element in boxes1 and boxes2
     """
 
-    def box_area(box):
-        # box = 4xn
-        return (box[2] - box[0]) * (box[3] - box[1])
-
-    area1 = box_area(box1.T)
-    area2 = box_area(box2.T)
-
     # inter(N,M) = (rb(N,M,2) - lt(N,M,2)).clamp(0).prod(2)
-    inter = (torch.min(box1[:, None, 2:], box2[:, 2:]) - torch.max(box1[:, None, :2], box2[:, :2])).clamp(0).prod(2)
-    return inter / (area1[:, None] + area2 - inter)  # iou = inter / (area1 + area2 - inter)
+    (a1, a2), (b1, b2) = box1[:, None].chunk(2, 2), box2.chunk(2, 1)
+    inter = (torch.min(a2, b2) - torch.max(a1, b1)).clamp(0).prod(2)
+
+    # IoU = inter / (area1 + area2 - inter)
+    return inter / (box_area(box1.T)[:, None] + box_area(box2.T) - inter)
 
 
 def bbox_ioa(box1, box2, eps=1E-7):
@@ -280,11 +280,9 @@ def bbox_ioa(box1, box2, eps=1E-7):
     returns:    np.array of shape(n)
     """
 
-    box2 = box2.transpose()
-
     # Get the coordinates of bounding boxes
-    b1_x1, b1_y1, b1_x2, b1_y2 = box1[0], box1[1], box1[2], box1[3]
-    b2_x1, b2_y1, b2_x2, b2_y2 = box2[0], box2[1], box2[2], box2[3]
+    b1_x1, b1_y1, b1_x2, b1_y2 = box1
+    b2_x1, b2_y1, b2_x2, b2_y2 = box2.T
 
     # Intersection area
     inter_area = (np.minimum(b1_x2, b2_x2) - np.maximum(b1_x1, b2_x1)).clip(0) * \

val.py

@@ -38,10 +38,10 @@ ROOT = Path(os.path.relpath(ROOT, Path.cwd()))  # relative
 from models.common import DetectMultiBackend
 from utils.callbacks import Callbacks
 from utils.datasets import create_dataloader
-from utils.general import (LOGGER, box_iou, check_dataset, check_img_size, check_requirements, check_yaml,
+from utils.general import (LOGGER, check_dataset, check_img_size, check_requirements, check_yaml,
                            coco80_to_coco91_class, colorstr, increment_path, non_max_suppression, print_args,
                            scale_coords, xywh2xyxy, xyxy2xywh)
-from utils.metrics import ConfusionMatrix, ap_per_class
+from utils.metrics import ConfusionMatrix, ap_per_class, box_iou
 from utils.plots import output_to_target, plot_images, plot_val_study
 from utils.torch_utils import select_device, time_sync