Upload 3 files

utils/general.py

@@ -0,0 +1,748 @@
+import glob
+import logging
+import math
+import random
+import re
+import time
+from pathlib import Path
+
+import cv2
+import numpy as np
+import torch
+import torch.nn.functional as F
+import torchvision
+from matplotlib import pyplot as plt
+
+id_dict = {'person': 0, 'rider': 1, 'car': 2, 'bus': 3, 'truck': 4,
+           'bike': 5, 'motor': 6, 'tl_green': 7, 'tl_red': 8,
+           'tl_yellow': 9, 'tl_none': 10, 'traffic sign': 11, 'train': 12}
+id_dict_single = {'car': 0, 'bus': 1, 'truck': 2, 'train': 3}
+
+
+def clean_str(s):
+    # Cleans a string by replacing special characters with underscore _
+    return re.sub(pattern="[|@#!¡·$€%&()=?¿^*;:,¨´><+]", repl="_", string=s)
+
+
+def set_logging(rank=-1):
+    logging.basicConfig(
+        format="%(message)s",
+        level=logging.INFO if rank in [-1, 0] else logging.WARN)
+
+
+def convert(size, box):
+    dw = 1. / (size[0])
+    dh = 1. / (size[1])
+    x = (box[0] + box[1]) / 2.0
+    y = (box[2] + box[3]) / 2.0
+    w = box[1] - box[0]
+    h = box[3] - box[2]
+    x = x * dw
+    w = w * dw
+    y = y * dh
+    h = h * dh
+    return (x, y, w, h)
+
+
+def xywh2xyxy(x):
+    # Convert nx4 boxes from [x, y, w, h] to [x1, y1, x2, y2] where xy1=top-left, xy2=bottom-right
+    y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x)
+    y[:, 0] = x[:, 0] - x[:, 2] / 2  # top left x
+    y[:, 1] = x[:, 1] - x[:, 3] / 2  # top left y
+    y[:, 2] = x[:, 0] + x[:, 2] / 2  # bottom right x
+    y[:, 3] = x[:, 1] + x[:, 3] / 2  # bottom right y
+    return y
+
+
+def xyxy2xywh(x):
+    # Convert nx4 boxes from [x1, y1, x2, y2] to [x, y, w, h] where xy1=top-left, xy2=bottom-right
+    y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x)
+    y[:, 0] = (x[:, 0] + x[:, 2]) / 2  # x center
+    y[:, 1] = (x[:, 1] + x[:, 3]) / 2  # y center
+    y[:, 2] = x[:, 2] - x[:, 0]  # width
+    y[:, 3] = x[:, 3] - x[:, 1]  # height
+    return y
+
+
+def bbox_iou(box1, box2, x1y1x2y2=True, GIoU=False, DIoU=False, CIoU=False, SIoU=False, WIoU=False, eps=1e-7):
+    # Returns the IoU of box1 to box2. box1 is 4, box2 is nx4
+    box2 = box2.T
+
+    # 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
+
+    # 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 = inter / union
+    if GIoU or DIoU or CIoU or SIoU:
+        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 SIoU:  # SIoU Loss https://arxiv.org/pdf/2205.12740.pdf
+            s_cw = (b2_x1 + b2_x2 - b1_x1 - b1_x2) * 0.5
+            s_ch = (b2_y1 + b2_y2 - b1_y1 - b1_y2) * 0.5
+            sigma = torch.pow(s_cw ** 2 + s_ch ** 2, 0.5)
+            sin_alpha_1 = torch.abs(s_cw) / sigma
+            sin_alpha_2 = torch.abs(s_ch) / sigma
+            threshold = pow(2, 0.5) / 2
+            sin_alpha = torch.where(sin_alpha_1 > threshold, sin_alpha_2, sin_alpha_1)
+            # angle_cost = 1 - 2 * torch.pow( torch.sin(torch.arcsin(sin_alpha) - np.pi/4), 2)
+            angle_cost = torch.cos(torch.arcsin(sin_alpha) * 2 - np.pi / 2)
+            rho_x = (s_cw / cw) ** 2
+            rho_y = (s_ch / ch) ** 2
+            gamma = angle_cost - 2
+            distance_cost = 2 - torch.exp(gamma * rho_x) - torch.exp(gamma * rho_y)
+            omiga_w = torch.abs(w1 - w2) / torch.max(w1, w2)
+            omiga_h = torch.abs(h1 - h2) / torch.max(h1, h2)
+            shape_cost = torch.pow(1 - torch.exp(-1 * omiga_w), 4) + torch.pow(1 - torch.exp(-1 * omiga_h), 4)
+            return iou - 0.5 * (distance_cost + shape_cost)
+        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
+            if DIoU:
+                return iou - rho2 / c2  # DIoU
+            elif 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():
+                    alpha = v / (v - iou + (1 + eps))
+                return iou - (rho2 / c2 + v * alpha)  # CIoU
+        else:  # GIoU https://arxiv.org/pdf/1902.09630.pdf
+            c_area = cw * ch + eps  # convex area
+            return iou - (c_area - union) / c_area  # GIoU
+    elif WIoU:
+        b1 = torch.stack([b1_x1, b1_y1, b1_x2, b1_y2], dim=-1)
+        b2 = torch.stack([b2_x1, b2_y1, b2_x2, b2_y2], dim=-1)
+
+        self = IoU_Cal(b1, b2)
+        loss = getattr(IoU_Cal, 'WIoU')(b1, b2, self=self)
+        iou = 1 - self.iou
+        return loss, iou
+    else:
+        return iou  # IoU
+
+
+def box_iou(box1, box2):
+    # https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py
+    """
+    Return intersection-over-union (Jaccard index) of boxes.
+    Both sets of boxes are expected to be in (x1, y1, x2, y2) format.
+    Arguments:
+        box1 (Tensor[N, 4])
+        box2 (Tensor[M, 4])
+    Returns:
+        iou (Tensor[N, M]): the NxM matrix containing the pairwise
+            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)
+
+
+def letterbox(combination, new_shape=(640, 640), color=(114, 114, 114), auto=True, scaleFill=False, scaleup=True):
+    """Resize the input image and automatically padding to suitable shape :https://zhuanlan.zhihu.com/p/172121380"""
+    # Resize image to a 32-pixel-multiple rectangle https://github.com/ultralytics/yolov3/issues/232
+    img, gray, line = combination
+    shape = img.shape[:2]  # current shape [height, width]
+    if isinstance(new_shape, int):
+        new_shape = (new_shape, new_shape)
+
+    # Scale ratio (new / old)
+    r = min(new_shape[0] / shape[0], new_shape[1] / shape[1])
+    if not scaleup:  # only scale down, do not scale up (for better test mAP)
+        r = min(r, 1.0)
+
+    # Compute padding
+    ratio = r, r  # width, height ratios
+    new_unpad = int(round(shape[1] * r)), int(round(shape[0] * r))
+    dw, dh = new_shape[1] - new_unpad[0], new_shape[0] - new_unpad[1]  # wh padding
+    # if auto:  # minimum rectangle
+    #     dw, dh = np.mod(dw, 32), np.mod(dh, 32)  # wh padding
+    # elif scaleFill:  # stretch
+    #     dw, dh = 0.0, 0.0
+    #     new_unpad = (new_shape[1], new_shape[0])
+    #     ratio = new_shape[1] / shape[1], new_shape[0] / shape[0]  # width, height ratios
+
+    dw /= 2  # divide padding into 2 sides
+    dh /= 2
+
+    if shape[::-1] != new_unpad:  # resize
+        img = cv2.resize(img, new_unpad, interpolation=cv2.INTER_LINEAR)
+        gray = cv2.resize(gray, new_unpad, interpolation=cv2.INTER_LINEAR)
+        line = cv2.resize(line, new_unpad, interpolation=cv2.INTER_LINEAR)
+
+    top, bottom = int(round(dh - 0.1)), int(round(dh + 0.1))
+    left, right = int(round(dw - 0.1)), int(round(dw + 0.1))
+
+    img = cv2.copyMakeBorder(img, top, bottom, left, right, cv2.BORDER_CONSTANT, value=color)  # add border
+    gray = cv2.copyMakeBorder(gray, top, bottom, left, right, cv2.BORDER_CONSTANT, value=0)  # add border
+    line = cv2.copyMakeBorder(line, top, bottom, left, right, cv2.BORDER_CONSTANT, value=0)  # add border
+    # print(img.shape)
+
+    combination = (img, gray, line)
+    return combination, ratio, (dw, dh)
+
+
+def letterbox_for_img(img, new_shape=(640, 640), color=(114, 114, 114), auto=True, scaleFill=False, scaleup=True):
+    # Resize image to a 32-pixel-multiple rectangle https://github.com/ultralytics/yolov3/issues/232
+    shape = img.shape[:2]  # current shape [height, width]
+    if isinstance(new_shape, int):
+        new_shape = (new_shape, new_shape)
+
+    # Scale ratio (new / old)
+    r = min(new_shape[0] / shape[0], new_shape[1] / shape[1])
+    if not scaleup:  # only scale down, do not scale up (for better test mAP)
+        r = min(r, 1.0)
+
+    # Compute padding
+    ratio = r, r  # width, height ratios
+    new_unpad = int(round(shape[1] * r)), int(round(shape[0] * r))
+    dw, dh = new_shape[1] - new_unpad[0], new_shape[0] - new_unpad[1]  # wh padding
+    if auto:  # minimum rectangle
+        dw, dh = np.mod(dw, 32), np.mod(dh, 32)  # wh padding
+    elif scaleFill:  # stretch
+        dw, dh = 0.0, 0.0
+        new_unpad = (new_shape[1], new_shape[0])
+        ratio = new_shape[1] / shape[1], new_shape[0] / shape[0]  # width, height ratios
+
+    dw /= 2  # divide padding into 2 sides
+    dh /= 2
+    if shape[::-1] != new_unpad:  # resize
+        img = cv2.resize(img, new_unpad, interpolation=cv2.INTER_AREA)
+
+    top, bottom = int(round(dh - 0.1)), int(round(dh + 0.1))
+    left, right = int(round(dw - 0.1)), int(round(dw + 0.1))
+    img = cv2.copyMakeBorder(img, top, bottom, left, right, cv2.BORDER_CONSTANT, value=color)  # add border
+    return img, ratio, (dw, dh)
+
+
+def colorstr(*input):
+    # Colors a string https://en.wikipedia.org/wiki/ANSI_escape_code, i.e.  colorstr('blue', 'hello world')
+    *args, string = input if len(input) > 1 else ('blue', 'bold', input[0])  # color arguments, string
+    colors = {'black': '\033[30m',  # basic colors
+              'red': '\033[31m',
+              'green': '\033[32m',
+              'yellow': '\033[33m',
+              'blue': '\033[34m',
+              'magenta': '\033[35m',
+              'cyan': '\033[36m',
+              'white': '\033[37m',
+              'bright_black': '\033[90m',  # bright colors
+              'bright_red': '\033[91m',
+              'bright_green': '\033[92m',
+              'bright_yellow': '\033[93m',
+              'bright_blue': '\033[94m',
+              'bright_magenta': '\033[95m',
+              'bright_cyan': '\033[96m',
+              'bright_white': '\033[97m',
+              'end': '\033[0m',  # misc
+              'bold': '\033[1m',
+              'underline': '\033[4m'}
+    return ''.join(colors[x] for x in args) + f'{string}' + colors['end']
+
+
+def make_divisible(x, divisor):
+    # Returns x evenly divisible by divisor
+    return math.ceil(x / divisor) * divisor
+
+
+def check_img_size(img_size, s=32):
+    # Verify img_size is a multiple of stride s
+    new_size = make_divisible(img_size, int(s))  # ceil gs-multiple
+    if new_size != img_size:
+        print('WARNING: --img-size %g must be multiple of max stride %g, updating to %g' % (img_size, s, new_size))
+    return new_size
+
+
+def scale_img(img, ratio=1.0, same_shape=False, gs=32):  # img(16,3,256,416)
+    # scales img(bs,3,y,x) by ratio constrained to gs-multiple
+    if ratio == 1.0:
+        return img
+    else:
+        h, w = img.shape[2:]
+        s = (int(h * ratio), int(w * ratio))  # new size
+        img = F.interpolate(img, size=s, mode='bilinear', align_corners=False)  # resize
+        if not same_shape:  # pad/crop img
+            h, w = [math.ceil(x * ratio / gs) * gs for x in (h, w)]
+        return F.pad(img, [0, w - s[1], 0, h - s[0]], value=0.447)  # value = imagenet mean
+
+
+def random_perspective(combination, targets=(), degrees=10, translate=.1, scale=.1, shear=10, perspective=0.0,
+                       border=(0, 0)):
+    """combination of img transform"""
+    # torchvision.transforms.RandomAffine(degrees=(-10, 10), translate=(.1, .1), scale=(.9, 1.1), shear=(-10, 10))
+    # targets = [cls, xyxy]
+    img, gray, line = combination
+    height = img.shape[0] + border[0] * 2  # shape(h,w,c)
+    width = img.shape[1] + border[1] * 2
+
+    # Center
+    C = np.eye(3)
+    C[0, 2] = -img.shape[1] / 2  # x translation (pixels)
+    C[1, 2] = -img.shape[0] / 2  # y translation (pixels)
+
+    # Perspective
+    P = np.eye(3)
+    P[2, 0] = random.uniform(-perspective, perspective)  # x perspective (about y)
+    P[2, 1] = random.uniform(-perspective, perspective)  # y perspective (about x)
+
+    # Rotation and Scale
+    R = np.eye(3)
+    a = random.uniform(-degrees, degrees)
+    # a += random.choice([-180, -90, 0, 90])  # add 90deg rotations to small rotations
+    s = random.uniform(1 - scale, 1 + scale)
+    # s = 2 ** random.uniform(-scale, scale)
+    R[:2] = cv2.getRotationMatrix2D(angle=a, center=(0, 0), scale=s)
+
+    # Shear
+    S = np.eye(3)
+    S[0, 1] = math.tan(random.uniform(-shear, shear) * math.pi / 180)  # x shear (deg)
+    S[1, 0] = math.tan(random.uniform(-shear, shear) * math.pi / 180)  # y shear (deg)
+
+    # Translation
+    T = np.eye(3)
+    T[0, 2] = random.uniform(0.5 - translate, 0.5 + translate) * width  # x translation (pixels)
+    T[1, 2] = random.uniform(0.5 - translate, 0.5 + translate) * height  # y translation (pixels)
+
+    # Combined rotation matrix
+    M = T @ S @ R @ P @ C  # order of operations (right to left) is IMPORTANT
+    if (border[0] != 0) or (border[1] != 0) or (M != np.eye(3)).any():  # image changed
+        if perspective:
+            img = cv2.warpPerspective(img, M, dsize=(width, height), borderValue=(114, 114, 114))
+            gray = cv2.warpPerspective(gray, M, dsize=(width, height), borderValue=0)
+            line = cv2.warpPerspective(line, M, dsize=(width, height), borderValue=0)
+        else:  # affine
+            img = cv2.warpAffine(img, M[:2], dsize=(width, height), borderValue=(114, 114, 114))
+            gray = cv2.warpAffine(gray, M[:2], dsize=(width, height), borderValue=0)
+            line = cv2.warpAffine(line, M[:2], dsize=(width, height), borderValue=0)
+
+    # Visualize
+    # import matplotlib.pyplot as plt
+    # ax = plt.subplots(1, 2, figsize=(12, 6))[1].ravel()
+    # ax[0].imshow(img[:, :, ::-1])  # base
+    # ax[1].imshow(img2[:, :, ::-1])  # warped
+
+    # Transform label coordinates
+    n = len(targets)
+    if n:
+        # warp points
+        xy = np.ones((n * 4, 3))
+        xy[:, :2] = targets[:, [1, 2, 3, 4, 1, 4, 3, 2]].reshape(n * 4, 2)  # x1y1, x2y2, x1y2, x2y1
+        xy = xy @ M.T  # transform
+        if perspective:
+            xy = (xy[:, :2] / xy[:, 2:3]).reshape(n, 8)  # rescale
+        else:  # affine
+            xy = xy[:, :2].reshape(n, 8)
+
+        # create new boxes
+        x = xy[:, [0, 2, 4, 6]]
+        y = xy[:, [1, 3, 5, 7]]
+        xy = np.concatenate((x.min(1), y.min(1), x.max(1), y.max(1))).reshape(4, n).T
+
+        # # apply angle-based reduction of bounding boxes
+        # radians = a * math.pi / 180
+        # reduction = max(abs(math.sin(radians)), abs(math.cos(radians))) ** 0.5
+        # x = (xy[:, 2] + xy[:, 0]) / 2
+        # y = (xy[:, 3] + xy[:, 1]) / 2
+        # w = (xy[:, 2] - xy[:, 0]) * reduction
+        # h = (xy[:, 3] - xy[:, 1]) * reduction
+        # xy = np.concatenate((x - w / 2, y - h / 2, x + w / 2, y + h / 2)).reshape(4, n).T
+
+        # clip boxes
+        xy[:, [0, 2]] = xy[:, [0, 2]].clip(0, width)
+        xy[:, [1, 3]] = xy[:, [1, 3]].clip(0, height)
+
+        # filter candidates
+        i = _box_candidates(box1=targets[:, 1:5].T * s, box2=xy.T)
+        targets = targets[i]
+        targets[:, 1:5] = xy[i]
+
+    combination = (img, gray, line)
+    return combination, targets
+
+
+def _box_candidates(box1, box2, wh_thr=2, ar_thr=20, area_thr=0.1):  # box1(4,n), box2(4,n)
+    # Compute candidate boxes: box1 before augment, box2 after augment, wh_thr (pixels), aspect_ratio_thr, area_ratio
+    w1, h1 = box1[2] - box1[0], box1[3] - box1[1]
+    w2, h2 = box2[2] - box2[0], box2[3] - box2[1]
+    ar = np.maximum(w2 / (h2 + 1e-16), h2 / (w2 + 1e-16))  # aspect ratio
+    return (w2 > wh_thr) & (h2 > wh_thr) & (w2 * h2 / (w1 * h1 + 1e-16) > area_thr) & (ar < ar_thr)  # candidates
+
+
+def mixup(im, labels, seg_label, lane_label, im2, labels2, seg_label2, lane_label2):
+    # Applies MixUp augmentation https://arxiv.org/pdf/1710.09412.pdf
+    r = np.random.beta(32.0, 32.0)  # mixup ratio, alpha=beta=32.0
+    im = (im * r + im2 * (1 - r)).astype(np.uint8)
+    labels = np.concatenate((labels, labels2), 0)
+    seg_label |= seg_label2
+    lane_label |= lane_label2
+    return im, labels, seg_label, lane_label
+
+
+def augment_hsv(img, hgain=0.5, sgain=0.5, vgain=0.5):
+    """change color hue, saturation, value"""
+    r = np.random.uniform(-1, 1, 3) * [hgain, sgain, vgain] + 1  # random gains
+    hue, sat, val = cv2.split(cv2.cvtColor(img, cv2.COLOR_BGR2HSV))
+    dtype = img.dtype  # uint8
+
+    x = np.arange(0, 256, dtype=np.int16)
+    lut_hue = ((x * r[0]) % 180).astype(dtype)
+    lut_sat = np.clip(x * r[1], 0, 255).astype(dtype)
+    lut_val = np.clip(x * r[2], 0, 255).astype(dtype)
+
+    img_hsv = cv2.merge((cv2.LUT(hue, lut_hue), cv2.LUT(sat, lut_sat), cv2.LUT(val, lut_val))).astype(dtype)
+    cv2.cvtColor(img_hsv, cv2.COLOR_HSV2BGR, dst=img)  # no return needed
+
+    # Histogram equalization
+    # if random.random() < 0.2:
+    #     for i in range(3):
+    #         img[:, :, i] = cv2.equalizeHist(img[:, :, i])
+
+
+def non_max_suppression(prediction, conf_thres=0.25, iou_thres=0.45, classes=None, agnostic=False, multi_label=False,
+                        labels=()):
+    """Runs Non-Maximum Suppression (NMS) on inference results
+
+    Returns:
+         list of detections, on (n,6) tensor per image [xyxy, conf, cls]
+    """
+
+    nc = prediction.shape[2] - 5  # number of classes
+    xc = prediction[..., 4] > conf_thres  # candidates
+
+    # Settings
+    min_wh, max_wh = 2, 4096  # (pixels) minimum and maximum box width and height
+    max_det = 300  # maximum number of detections per image
+    max_nms = 30000  # maximum number of boxes into torchvision.ops.nms()
+    time_limit = 10.0  # seconds to quit after
+    redundant = True  # require redundant detections
+    multi_label &= nc > 1  # multiple labels per box (adds 0.5ms/img)
+    merge = False  # use merge-NMS
+
+    t = time.time()
+    output = [torch.zeros((0, 6), device=prediction.device)] * prediction.shape[0]
+    for xi, x in enumerate(prediction):  # image index, image inference
+        # Apply constraints
+        # x[((x[..., 2:4] < min_wh) | (x[..., 2:4] > max_wh)).any(1), 4] = 0  # width-height
+        x = x[xc[xi]]  # confidence
+
+        # Cat apriori labels if autolabelling
+        if labels and len(labels[xi]):
+            l = labels[xi]
+            v = torch.zeros((len(l), nc + 5), device=x.device)
+            v[:, :4] = l[:, 1:5]  # box
+            v[:, 4] = 1.0  # conf
+            v[range(len(l)), l[:, 0].long() + 5] = 1.0  # cls
+            x = torch.cat((x, v), 0)
+
+        # If none remain process next image
+        if not x.shape[0]:
+            continue
+
+        # Compute conf
+        if nc == 1:
+            x[:, 5:] = x[:, 4:5]  # for models with one class, cls_loss is 0 and cls_conf is always 0.5,
+            # so there is no need to multiplicate.
+        else:
+            x[:, 5:] *= x[:, 4:5]  # conf = obj_conf * cls_conf
+
+        # Box (center x, center y, width, height) to (x1, y1, x2, y2)
+        box = xywh2xyxy(x[:, :4])
+
+        # Detections matrix nx6 (xyxy, conf, cls)
+        if multi_label:
+            i, j = (x[:, 5:] > conf_thres).nonzero(as_tuple=False).T
+            x = torch.cat((box[i], x[i, j + 5, None], j[:, None].float()), 1)
+        else:  # best class only
+            conf, j = x[:, 5:].max(1, keepdim=True)
+            x = torch.cat((box, conf, j.float()), 1)[conf.view(-1) > conf_thres]
+
+        # Filter by class
+        if classes is not None:
+            x = x[(x[:, 5:6] == torch.tensor(classes, device=x.device)).any(1)]
+
+        # Apply finite constraint
+        # if not torch.isfinite(x).all():
+        #     x = x[torch.isfinite(x).all(1)]
+
+        # Check shape
+        n = x.shape[0]  # number of boxes
+        if not n:  # no boxes
+            continue
+        elif n > max_nms:  # excess boxes
+            x = x[x[:, 4].argsort(descending=True)[:max_nms]]  # sort by confidence
+
+        # Batched NMS
+        c = x[:, 5:6] * (0 if agnostic else max_wh)  # classes
+        boxes, scores = x[:, :4] + c, x[:, 4]  # boxes (offset by class), scores
+        i = torchvision.ops.nms(boxes, scores, iou_thres)  # NMS
+        if i.shape[0] > max_det:  # limit detections
+            i = i[:max_det]
+        if merge and (1 < n < 3E3):  # Merge NMS (boxes merged using weighted mean)
+            # update boxes as boxes(i,4) = weights(i,n) * boxes(n,4)
+            iou = box_iou(boxes[i], boxes) > iou_thres  # iou matrix
+            weights = iou * scores[None]  # box weights
+            x[i, :4] = torch.mm(weights, x[:, :4]).float() / weights.sum(1, keepdim=True)  # merged boxes
+            if redundant:
+                i = i[iou.sum(1) > 1]  # require redundancy
+
+        output[xi] = x[i]
+        if (time.time() - t) > time_limit:
+            print(f'WARNING: NMS time limit {time_limit}s exceeded')
+            break  # time limit exceeded
+
+    return output
+
+
+def scale_coords(img1_shape, coords, img0_shape, ratio_pad=None):
+    # Rescale coords (xyxy) from img1_shape to img0_shape
+    if ratio_pad is None:  # calculate from img0_shape
+        gain = min(img1_shape[0] / img0_shape[0], img1_shape[1] / img0_shape[1])  # gain  = old / new
+        pad = (img1_shape[1] - img0_shape[1] * gain) / 2, (img1_shape[0] - img0_shape[0] * gain) / 2  # wh padding
+    else:
+        gain = ratio_pad[0][0]
+        pad = ratio_pad[1]
+
+    coords[:, [0, 2]] -= pad[0]  # x padding
+    coords[:, [1, 3]] -= pad[1]  # y padding
+    coords[:, :4] /= gain
+    clip_coords(coords, img0_shape)
+    return coords
+
+
+def clip_coords(boxes, img_shape):
+    # Clip bounding xyxy bounding boxes to image shape (height, width)
+    boxes[:, 0].clamp_(0, img_shape[1])  # x1
+    boxes[:, 1].clamp_(0, img_shape[0])  # y1
+    boxes[:, 2].clamp_(0, img_shape[1])  # x2
+    boxes[:, 3].clamp_(0, img_shape[0])  # y2
+
+
+def ap_per_class(tp, conf, pred_cls, target_cls, plot=False, save_dir='precision-recall_curve.png', names=[]):
+    """ Compute the average precision, given the recall and precision curves.
+    Source: https://github.com/rafaelpadilla/Object-Detection-Metrics.
+    # Arguments
+        tp:  True positives (nparray, nx1 or nx10).
+        conf:  Objectness value from 0-1 (nparray).
+        pred_cls:  Predicted object classes (nparray).
+        target_cls:  True object classes (nparray).
+        plot:  Plot precision-recall curve at mAP@0.5
+        save_dir:  Plot save directory
+    # Returns
+        The average precision as computed in py-faster-rcnn.
+    """
+
+    # Sort by objectness
+    i = np.argsort(-conf)
+    tp, conf, pred_cls = tp[i], conf[i], pred_cls[i]
+
+    # Find unique classes
+    unique_classes = np.unique(target_cls)
+
+    # Create Precision-Recall curve and compute AP for each class
+    px, py = np.linspace(0, 1, 1000), []  # for plotting
+    pr_score = 0.1  # score to evaluate P and R https://github.com/ultralytics/yolov3/issues/898
+    s = [unique_classes.shape[0], tp.shape[1]]  # number class, number iou thresholds (i.e. 10 for mAP0.5...0.95)
+    ap, p, r = np.zeros(s), np.zeros((unique_classes.shape[0], 1000)), np.zeros((unique_classes.shape[0], 1000))
+    for ci, c in enumerate(unique_classes):
+        i = pred_cls == c
+        n_l = (target_cls == c).sum()  # number of labels
+        n_p = i.sum()  # number of predictions
+
+        if n_p == 0 or n_l == 0:
+            continue
+        else:
+            # Accumulate FPs and TPs
+            fpc = (1 - tp[i]).cumsum(0)
+            tpc = tp[i].cumsum(0)
+
+            # Recall
+            recall = tpc / (n_l + 1e-16)  # recall curve
+            r[ci] = np.interp(-px, -conf[i], recall[:, 0], left=0)  # negative x, xp because xp decreases
+
+            # Precision
+            precision = tpc / (tpc + fpc)  # precision curve
+            p[ci] = np.interp(-px, -conf[i], precision[:, 0], left=1)  # p at pr_score
+            # AP from recall-precision curve
+            for j in range(tp.shape[1]):
+                ap[ci, j], mpre, mrec = compute_ap(recall[:, j], precision[:, j])
+                if plot and (j == 0):
+                    py.append(np.interp(px, mrec, mpre))  # precision at mAP@0.5
+
+    # Compute F1 score (harmonic mean of precision and recall)
+    f1 = 2 * p * r / (p + r + 1e-16)
+    i = r.mean(0).argmax()
+
+    if plot:
+        plot_pr_curve(px, py, ap, save_dir, names)
+
+    return p[:, i], r[:, i], ap, f1[:, i], unique_classes.astype('int32')
+
+
+def compute_ap(recall, precision):
+    """ Compute the average precision, given the recall and precision curves.
+    Source: https://github.com/rbgirshick/py-faster-rcnn.
+    # Arguments
+        recall:    The recall curve (list).
+        precision: The precision curve (list).
+    # Returns
+        The average precision as computed in py-faster-rcnn.
+    """
+
+    # Append sentinel values to beginning and end
+    mrec = np.concatenate(([0.], recall, [recall[-1] + 1E-3]))
+    mpre = np.concatenate(([1.], precision, [0.]))
+
+    # Compute the precision envelope
+    mpre = np.flip(np.maximum.accumulate(np.flip(mpre)))
+
+    # Integrate area under curve
+    method = 'interp'  # methods: 'continuous', 'interp'
+    if method == 'interp':
+        x = np.linspace(0, 1, 101)  # 101-point interp (COCO)
+        ap = np.trapz(np.interp(x, mrec, mpre), x)  # integrate
+
+    else:  # 'continuous'
+        i = np.where(mrec[1:] != mrec[:-1])[0]  # points where x axis (recall) changes
+        ap = np.sum((mrec[i + 1] - mrec[i]) * mpre[i + 1])  # area under curve
+
+    return ap, mpre, mrec
+
+
+def plot_pr_curve(px, py, ap, save_dir='.', names=()):
+    fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True)
+    py = np.stack(py, axis=1)
+
+    if 0 < len(names) < 21:  # show mAP in legend if < 10 classes
+        for i, y in enumerate(py.T):
+            ax.plot(px, y, linewidth=1, label=f'{names[i]} %.3f' % ap[i, 0])  # plot(recall, precision)
+    else:
+        ax.plot(px, py, linewidth=1, color='grey')  # plot(recall, precision)
+
+    ax.plot(px, py.mean(1), linewidth=3, color='blue', label='all classes %.3f mAP@0.5' % ap[:, 0].mean())
+    ax.set_xlabel('Recall')
+    ax.set_ylabel('Precision')
+    ax.set_xlim(0, 1)
+    ax.set_ylim(0, 1)
+    plt.legend(bbox_to_anchor=(1.04, 1), loc="upper left")
+    fig.savefig(Path(save_dir) / 'precision_recall_curve.png', dpi=250)
+
+
+def increment_path(path, exist_ok=True, sep=''):
+    # Increment path, i.e. runs/exp --> runs/exp{sep}0, runs/exp{sep}1 etc.
+    path = Path(path)  # os-agnostic
+    if (path.exists() and exist_ok) or (not path.exists()):
+        return str(path)
+    else:
+        dirs = glob.glob(f"{path}{sep}*")  # similar paths
+        matches = [re.search(rf"%s{sep}(\d+)" % path.stem, d) for d in dirs]
+        i = [int(m.groups()[0]) for m in matches if m]  # indices
+        n = max(i) + 1 if i else 2  # increment number
+        return f"{path}{sep}{n}"  # update path
+
+
+class IoU_Cal:
+    ''' pred, target: x0,y0,x1,y1
+        monotonous: {
+            None: origin
+            True: monotonic FM
+            False: non-monotonic FM
+        }
+        momentum: The momentum of running mean'''
+    iou_mean = 1.
+    monotonous = False
+    momentum = 1 - 0.5 ** (1 / 7000)
+
+    _is_train = True
+
+    def __init__(self, pred, target):
+        self.pred, self.target = pred, target
+        self._fget = {
+            # x,y,w,h
+            'pred_xy': lambda: (self.pred[..., :2] + self.pred[..., 2: 4]) / 2,
+            'pred_wh': lambda: self.pred[..., 2: 4] - self.pred[..., :2],
+            'target_xy': lambda: (self.target[..., :2] + self.target[..., 2: 4]) / 2,
+            'target_wh': lambda: self.target[..., 2: 4] - self.target[..., :2],
+            # x0,y0,x1,y1
+            'min_coord': lambda: torch.minimum(self.pred[..., :4], self.target[..., :4]),
+            'max_coord': lambda: torch.maximum(self.pred[..., :4], self.target[..., :4]),
+            # The overlapping region
+            'wh_inter': lambda: self.min_coord[..., 2: 4] - self.max_coord[..., :2],
+            's_inter': lambda: torch.prod(torch.relu(self.wh_inter), dim=-1),
+            # The area covered
+            's_union': lambda: torch.prod(self.pred_wh, dim=-1) +
+                               torch.prod(self.target_wh, dim=-1) - self.s_inter,
+            # The smallest enclosing box
+            'wh_box': lambda: self.max_coord[..., 2: 4] - self.min_coord[..., :2],
+            's_box': lambda: torch.prod(self.wh_box, dim=-1),
+            'l2_box': lambda: torch.square(self.wh_box).sum(dim=-1),
+            # The central points' connection of the bounding boxes
+            'd_center': lambda: self.pred_xy - self.target_xy,
+            'l2_center': lambda: torch.square(self.d_center).sum(dim=-1),
+            # IoU
+            'iou': lambda: 1 - self.s_inter / self.s_union
+        }
+        self._update(self)
+
+    def __setitem__(self, key, value):
+        self._fget[key] = value
+
+    def __getattr__(self, item):
+        if callable(self._fget[item]):
+            self._fget[item] = self._fget[item]()
+        return self._fget[item]
+
+    @classmethod
+    def train(cls):
+        cls._is_train = True
+
+    @classmethod
+    def eval(cls):
+        cls._is_train = False
+
+    @classmethod
+    def _update(cls, self):
+        if cls._is_train: cls.iou_mean = (1 - cls.momentum) * cls.iou_mean + \
+                                         cls.momentum * self.iou.detach().mean().item()
+
+    def _scaled_loss(self, loss, gamma=1.9, delta=3):
+        if isinstance(self.monotonous, bool):
+            if self.monotonous:
+                loss *= (self.iou.detach() / self.iou_mean).sqrt()
+            else:
+                beta = self.iou.detach() / self.iou_mean
+                alpha = delta * torch.pow(gamma, beta - delta)
+                loss *= beta / alpha
+        return loss
+
+    @classmethod
+    def IoU(cls, pred, target, self=None):
+        self = self if self else cls(pred, target)
+        return self.iou
+
+    @classmethod
+    def WIoU(cls, pred, target, self=None):
+        self = self if self else cls(pred, target)
+        dist = torch.exp(self.l2_center / self.l2_box.detach())
+        return self._scaled_loss(dist * self.iou)

utils/plots.py

@@ -0,0 +1,89 @@
+import random
+
+import cv2
+import matplotlib
+import numpy as np
+import torch
+
+
+def show_seg_result(img, result, index=0, epoch=0, batch=0, save_dir=None, is_ll=False, palette=None, is_demo=False,
+                    is_gt=False):
+    if palette is None:
+        palette = np.random.randint(0, 255, size=(3, 3))
+    palette[0] = [0, 0, 0]
+    palette[1] = [0, 255, 0]
+    palette[2] = [255, 0, 0]
+    palette = np.array(palette)
+    assert palette.shape[0] == 3  # len(classes)
+    assert palette.shape[1] == 3
+    assert len(palette.shape) == 2
+
+    if not is_demo:
+        color_seg = np.zeros((result.shape[0], result.shape[1], 3), dtype=np.uint8)
+        for label, color in enumerate(palette):
+            color_seg[result == label, :] = color
+    else:
+        color_area = np.zeros((result[0].shape[0], result[0].shape[1], 3), dtype=np.uint8)
+
+        color_area[result[0] == 1] = [0, 255, 0]
+        color_area[result[1] == 1] = [255, 0, 0]
+        color_seg = color_area
+
+    # convert to BGR
+    color_seg = color_seg[..., ::-1]
+    # print(color_seg.shape)
+    color_mask = np.mean(color_seg, 2)
+    img[color_mask != 0] = img[color_mask != 0] * 0.5 + color_seg[color_mask != 0] * 0.5
+    # img = img * 0.5 + color_seg * 0.5
+    img = img.astype(np.uint8)
+    img = cv2.resize(img, (1280, 720), interpolation=cv2.INTER_LINEAR)
+
+    if not is_demo:
+        if not is_gt:
+            if not is_ll:
+                cv2.imwrite(save_dir + "/batch_{}_{}_{}_da_segresult.png".format(epoch, batch, index), img)
+            else:
+                cv2.imwrite(save_dir + "/batch_{}_{}_{}_ll_segresult.png".format(epoch, batch, index), img)
+        else:
+            if not is_ll:
+                cv2.imwrite(save_dir + "/batch_{}_{}_{}_da_seg_gt.png".format(epoch, batch, index), img)
+            else:
+                cv2.imwrite(save_dir + "/batch_{}_{}_{}_ll_seg_gt.png".format(epoch, batch, index), img)
+    return img
+
+
+def plot_one_box(x, img, color=None, label=None, line_thickness=3):
+    # Plots one bounding box on image img
+    tl = line_thickness or round(0.002 * (img.shape[0] + img.shape[1]) / 2) + 1  # line/font thickness
+    color = color or [random.randint(0, 255) for _ in range(3)]
+    c1, c2 = (int(x[0]), int(x[1])), (int(x[2]), int(x[3]))
+    cv2.rectangle(img, c1, c2, color, thickness=tl, lineType=cv2.LINE_AA)
+    if label:
+        tf = max(tl - 1, 1)  # font thickness
+        t_size = cv2.getTextSize(label, 0, fontScale=tl / 3, thickness=tf)[0]
+        c2 = c1[0] + t_size[0], c1[1] - t_size[1] - 3
+        cv2.rectangle(img, c1, c2, color, -1, cv2.LINE_AA)  # filled
+        cv2.putText(img, label, (c1[0], c1[1] - 2), 0, tl / 3, [225, 255, 255], thickness=tf, lineType=cv2.LINE_AA)
+
+
+def driving_area_mask(da_seg_out, width, height, pad_w, pad_h, ratio):
+    da_predict = da_seg_out[:, :, pad_h:(height - pad_h), pad_w:(width - pad_w)]
+    da_seg_mask = torch.nn.functional.interpolate(da_predict, scale_factor=int(1 / ratio), mode='bilinear')
+    _, da_seg_mask = torch.max(da_seg_mask, 1)
+    da_seg_mask = da_seg_mask.int().squeeze().cpu().numpy()
+    return da_seg_mask
+
+
+def lane_line_mask(ll_seg_out, width, height, pad_w, pad_h, ratio):
+    ll_predict = ll_seg_out[:, :, pad_h:(height - pad_h), pad_w:(width - pad_w)]
+    ll_seg_mask = torch.nn.functional.interpolate(ll_predict, scale_factor=int(1 / ratio), mode='bilinear')
+    _, ll_seg_mask = torch.max(ll_seg_mask, 1)
+    ll_seg_mask = ll_seg_mask.int().squeeze().cpu().numpy()
+    return ll_seg_mask
+
+def color_list():
+    # Return first 10 plt colors as (r,g,b) https://stackoverflow.com/questions/51350872/python-from-color-name-to-rgb
+    def hex2rgb(h):
+        return tuple(int(h[1 + i:1 + i + 2], 16) for i in (0, 2, 4))
+
+    return [hex2rgb(h) for h in matplotlib.colors.TABLEAU_COLORS.values()]  # or BASE_ (8), CSS4_ (148), XKCD_ (949)

utils/torch_utils.py

@@ -0,0 +1,321 @@
+import datetime
+import logging
+import math
+import os
+import platform
+import subprocess
+import time
+from copy import deepcopy
+from pathlib import Path
+
+import thop
+import torch
+import torch.optim as optim
+from prefetch_generator import BackgroundGenerator
+from torch import nn
+from torch.utils.data import DataLoader
+
+logger = logging.getLogger(__name__)
+
+
+def create_logger(args, setting_yaml, phase='train', rank=-1):
+    # set up logger dir
+    dataset = setting_yaml['dataset_dataset']
+    model = setting_yaml['model_name']
+    cfg_path = os.path.basename(args.log_dir).split('.')[0]
+
+    if rank in [-1, 0]:
+        time_str = time.strftime('%Y-%m-%d-%H-%M')
+        log_file = '{}_{}_{}.log'.format(cfg_path, time_str, phase)
+        # set up tensorboard_log_dir
+        tensorboard_log_dir = Path(args.log_dir) / dataset / model / (cfg_path + '_' + time_str)
+        final_output_dir = tensorboard_log_dir
+        if not tensorboard_log_dir.exists():
+            print('=> creating {}'.format(tensorboard_log_dir))
+            tensorboard_log_dir.mkdir(parents=True)
+
+        final_log_file = tensorboard_log_dir / log_file
+        head = '%(asctime)-15s %(message)s'
+        logging.basicConfig(filename=str(final_log_file),
+                            format=head)
+        logger = logging.getLogger()
+        logger.setLevel(logging.INFO)
+        console = logging.StreamHandler()
+        logging.getLogger('').addHandler(console)
+
+        return logger, str(final_output_dir), str(tensorboard_log_dir)
+    else:
+        return None, None, None
+
+
+def select_device(logger=None, device='', batch_size=None):
+    # device = 'cpu' or '0' or '0,1,2,3'
+    s = f'mtpnet 🚀 {git_describe() or date_modified()} torch {torch.__version__} '  # string
+    cpu = device.lower() == 'cpu'
+    if cpu:
+        os.environ['CUDA_VISIBLE_DEVICES'] = '-1'  # force torch.cuda.is_available() = False
+    elif device:  # non-cpu device requested
+        os.environ['CUDA_VISIBLE_DEVICES'] = device  # set environment variable
+        assert torch.cuda.is_available(), f'CUDA unavailable, invalid device {device} requested'  # check availability
+
+    cuda = not cpu and torch.cuda.is_available()
+    if cuda:
+        n = torch.cuda.device_count()
+        if n > 1 and batch_size:  # check that batch_size is compatible with device_count
+            assert batch_size % n == 0, f'batch-size {batch_size} not multiple of GPU count {n}'
+        space = ' ' * len(s)
+        for i, d in enumerate(device.split(',') if device else range(n)):
+            p = torch.cuda.get_device_properties(i)
+            s += f"{'' if i == 0 else space}CUDA:{d} ({p.name}, {p.total_memory / 1024 ** 2}MB)\n"  # bytes to MB
+    else:
+        s += 'CPU\n'
+
+    if logger:
+        logger.info(s.encode().decode('ascii', 'ignore') if platform.system() == 'Windows' else s)  # emoji-safe
+    return torch.device('cuda:0' if cuda else 'cpu')
+
+
+def get_optimizer(setting_yaml, model):
+    optimizer = None
+    if setting_yaml['train_optimizer'] == 'sgd':
+        optimizer = optim.SGD(
+            filter(lambda p: p.requires_grad, model.parameters()),
+            lr=setting_yaml['train_lr0'],
+            momentum=setting_yaml['train_momentum'],
+            weight_decay=setting_yaml['train_wd'],
+            nesterov=setting_yaml['train_nesterov']
+        )
+    elif setting_yaml['train_optimizer'] == 'adam':
+        optimizer = optim.Adam(
+            filter(lambda p: p.requires_grad, model.parameters()),
+            # model.parameters(),
+            lr=setting_yaml['train_lr0'],
+            betas=(setting_yaml['train_momentum'], 0.999),
+            eps=1e-5
+        )
+
+    return optimizer
+
+
+def save_checkpoint(epoch, name, model, optimizer, scheduler, ema, output_dir, is_best, best_fitness):
+    model_state = model.module.state_dict() if is_parallel(model) else model.state_dict()
+    checkpoint = {
+        'epoch': epoch,
+        'model': name,
+        'state_dict': model_state,
+        'ema': deepcopy(ema.ema).half(),
+        'updates': ema.updates,
+        'optimizer': optimizer.state_dict(),
+        'scheduler': scheduler.state_dict(),
+        'best_fitness': best_fitness
+    }
+    last = os.path.join(output_dir, 'last.pth')
+    last_st = os.path.join(output_dir, 'last_st.pth')
+    best = os.path.join(output_dir, 'best.pth')
+    best_st = os.path.join(output_dir, 'best_st.pth')
+    torch.save(checkpoint, last)
+    torch.save(model_state, last_st)
+    if is_best and (epoch <= 100):
+        torch.save(checkpoint, best)
+        torch.save(model_state, best_st)
+    if is_best and (epoch > 100):
+        torch.save(checkpoint, os.path.join(output_dir, 'best_{:03d}.pth'.format(epoch)))
+        torch.save(model_state, os.path.join(output_dir, 'best_st_{:03d}.pth'.format(epoch)))
+
+
+class ModelEMA:
+    """ Model Exponential Moving Average from https://github.com/rwightman/pytorch-image-models
+    Keep a moving average of everything in the model state_dict (parameters and buffers).
+    This is intended to allow functionality like
+    https://www.tensorflow.org/api_docs/python/tf/train/ExponentialMovingAverage
+    A smoothed version of the weights is necessary for some training schemes to perform well.
+    This class is sensitive where it is initialized in the sequence of model init,
+    GPU assignment and distributed training wrappers.
+    """
+
+    def __init__(self, model, decay=0.9999, updates=0):
+        # Create EMA
+        self.ema = deepcopy(model.module if is_parallel(model) else model).eval()  # FP32 EMA
+        # if next(model.parameters()).device.type != 'cpu':
+        #     self.ema.half()  # FP16 EMA
+        self.updates = updates  # number of EMA updates
+        self.decay = lambda x: decay * (1 - math.exp(-x / 2000))  # decay exponential ramp (to help early epochs)
+        for p in self.ema.parameters():
+            p.requires_grad_(False)
+
+    def update(self, model):
+        # Update EMA parameters
+        with torch.no_grad():
+            self.updates += 1
+            d = self.decay(self.updates)
+
+            msd = model.module.state_dict() if is_parallel(model) else model.state_dict()  # model state_dict
+            for k, v in self.ema.state_dict().items():
+                if v.dtype.is_floating_point:
+                    v *= d
+                    v += (1. - d) * msd[k].detach()
+
+    def update_attr(self, model, include=(), exclude=('process_group', 'reducer')):
+        # Update EMA attributes
+        copy_attr(self.ema, model, include, exclude)
+
+
+class DataLoaderX(DataLoader):
+    """prefetch dataloader"""
+    def __iter__(self):
+        return BackgroundGenerator(super().__iter__())
+
+
+class AverageMeter(object):
+    """Computes and stores the average and current value"""
+
+    def __init__(self):
+        self.val = 0
+        self.avg = 0
+        self.sum = 0
+        self.count = 0
+
+    def reset(self):
+        self.val = 0
+        self.avg = 0
+        self.sum = 0
+        self.count = 0
+
+    def update(self, val, n=1):
+        self.val = val
+        self.sum += val * n
+        self.count += n
+        self.avg = self.sum / self.count if self.count != 0 else 0
+
+
+def git_describe(path=Path(__file__).parent):  # path must be a directory
+    # return human-readable git description, i.e. v5.0-5-g3e25f1e https://git-scm.com/docs/git-describe
+    s = f'git -C {path} describe --tags --long --always'
+    try:
+        return subprocess.check_output(s, shell=True, stderr=subprocess.STDOUT).decode()[:-1]
+    except subprocess.CalledProcessError as e:
+        return ''  # not a git repository
+
+
+def date_modified(path=__file__):
+    # return human-readable file modification date, i.e. '2021-3-26'
+    t = datetime.datetime.fromtimestamp(Path(path).stat().st_mtime)
+    return f'{t.year}-{t.month}-{t.day}'
+
+
+def is_parallel(model):
+    return type(model) in (nn.parallel.DataParallel, nn.parallel.DistributedDataParallel)
+
+
+def copy_attr(a, b, include=(), exclude=()):
+    # Copy attributes from b to a, options to only include [...] and to exclude [...]
+    for k, v in b.__dict__.items():
+        if (len(include) and k not in include) or k.startswith('_') or k in exclude:
+            continue
+        else:
+            setattr(a, k, v)
+
+
+def time_synchronized():
+    # pytorch-accurate time
+    if torch.cuda.is_available():
+        torch.cuda.synchronize()
+    return time.time()
+
+
+def profile(x, ops, n=100, device=None):
+    # profile a pytorch module or list of modules. Example usage:
+    #     x = torch.randn(16, 3, 640, 640)  # input
+    #     m1 = lambda x: x * torch.sigmoid(x)
+    #     m2 = nn.SiLU()
+    #     profile(x, [m1, m2], n=100)  # profile speed over 100 iterations
+
+    device = device or torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
+    x = x.to(device)
+    x.requires_grad = True
+    print(torch.__version__, device.type, torch.cuda.get_device_properties(0) if device.type == 'cuda' else '')
+    print(f"\n{'Params':>12s}{'GFLOPS':>12s}{'forward (ms)':>16s}{'backward (ms)':>16s}{'input':>24s}{'output':>24s}")
+    for m in ops if isinstance(ops, list) else [ops]:
+        m = m.to(device) if hasattr(m, 'to') else m  # device
+        m = m.half() if hasattr(m, 'half') and isinstance(x, torch.Tensor) and x.dtype is torch.float16 else m  # type
+        dtf, dtb, t = 0., 0., [0., 0., 0.]  # dt forward, backward
+        try:
+            flops = thop.profile(m, inputs=(x,), verbose=False)[0] / 1E9 * 2  # GFLOPS
+        except:
+            flops = 0
+
+        for _ in range(n):
+            t[0] = time_synchronized()
+            y = m(x)
+            t[1] = time_synchronized()
+            try:
+                _ = y.sum().backward()
+                t[2] = time_synchronized()
+            except:  # no backward method
+                t[2] = float('nan')
+            dtf += (t[1] - t[0]) * 1000 / n  # ms per op forward
+            dtb += (t[2] - t[1]) * 1000 / n  # ms per op backward
+
+        s_in = tuple(x.shape) if isinstance(x, torch.Tensor) else 'list'
+        s_out = tuple(y.shape) if isinstance(y, torch.Tensor) else 'list'
+        p = sum(list(x.numel() for x in m.parameters())) if isinstance(m, nn.Module) else 0  # parameters
+        print(f'{p:12}{flops:12.4g}{dtf:16.4g}{dtb:16.4g}{str(s_in):>24s}{str(s_out):>24s}')
+
+
+def initialize_weights(model):
+    for m in model.modules():
+        t = type(m)
+        if t is nn.Conv2d:
+            pass  # nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
+        elif t is nn.BatchNorm2d:
+            m.eps = 1e-3
+            m.momentum = 0.03
+        elif t in [nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6]:
+            m.inplace = True
+
+
+def fuse_conv_and_bn(conv, bn):
+    # Fuse convolution and batchnorm layers https://tehnokv.com/posts/fusing-batchnorm-and-conv/
+    fusedconv = nn.Conv2d(conv.in_channels,
+                          conv.out_channels,
+                          kernel_size=conv.kernel_size,
+                          stride=conv.stride,
+                          padding=conv.padding,
+                          groups=conv.groups,
+                          bias=True).requires_grad_(False).to(conv.weight.device)
+
+    # prepare filters
+    w_conv = conv.weight.clone().view(conv.out_channels, -1)
+    w_bn = torch.diag(bn.weight.div(torch.sqrt(bn.eps + bn.running_var)))
+    fusedconv.weight.copy_(torch.mm(w_bn, w_conv).view(fusedconv.weight.shape))
+
+    # prepare spatial bias
+    b_conv = torch.zeros(conv.weight.size(0), device=conv.weight.device) if conv.bias is None else conv.bias
+    b_bn = bn.bias - bn.weight.mul(bn.running_mean).div(torch.sqrt(bn.running_var + bn.eps))
+    fusedconv.bias.copy_(torch.mm(w_bn, b_conv.reshape(-1, 1)).reshape(-1) + b_bn)
+
+    return fusedconv
+
+
+def model_info(model, verbose=False, img_size=640):
+    # Model information. img_size may be int or list, i.e. img_size=640 or img_size=[640, 320]
+    n_p = sum(x.numel() for x in model.parameters())  # number parameters
+    n_g = sum(x.numel() for x in model.parameters() if x.requires_grad)  # number gradients
+    if verbose:
+        print('%5s %40s %9s %12s %20s %10s %10s' % ('layer', 'name', 'gradient', 'parameters', 'shape', 'mu', 'sigma'))
+        for i, (name, p) in enumerate(model.named_parameters()):
+            name = name.replace('module_list.', '')
+            print('%5g %40s %9s %12g %20s %10.3g %10.3g' %
+                  (i, name, p.requires_grad, p.numel(), list(p.shape), p.mean(), p.std()))
+
+    try:  # FLOPS
+        from thop import profile
+        stride = max(int(model.stride.max()), 32) if hasattr(model, 'stride') else 32
+        img = torch.zeros((1, model.yaml.get('ch', 3), stride, stride), device=next(model.parameters()).device)  # input
+        flops = profile(deepcopy(model), inputs=(img,), verbose=False)[0] / 1E9 * 2  # stride GFLOPS
+        img_size = img_size if isinstance(img_size, list) else [img_size, img_size]  # expand if int/float
+        fs = ', %.1f GFLOPS' % (flops * img_size[0] / stride * img_size[1] / stride)  # 640x640 GFLOPS
+    except (ImportError, Exception):
+        fs = ''
+
+    logger.info(f"Model Summary: {len(list(model.modules()))} layers, {n_p} parameters, {n_g} gradients{fs}")