Trainer + Dataloaders (#27)

Co-authored-by: Laughing-q <1185102784@qq.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: Ayush Chaurasia <ayushchaurasia@Ayushs-MacBook-Pro.local>
Co-authored-by: Ayush Chaurasia <ayush.chuararsia@gmail.com>

ultralytics/yolo/data/__init__.py

@@ -0,0 +1,3 @@
+from .build import build_classification_dataloader, build_dataloader
+from .dataset import ClassificationDataset, SemanticDataset, YOLODataset
+from .dataset_wrappers import MixAndRectDataset

ultralytics/yolo/data/augment.py

@@ -0,0 +1,785 @@
+import collections
+import math
+import random
+from copy import deepcopy
+
+import cv2
+import numpy as np
+import torch
+import torchvision.transforms as T
+
+from ..utils.general import LOGGER, check_version, colorstr, segment2box
+from ..utils.instance import Instances
+from ..utils.metrics import bbox_ioa
+from .utils import IMAGENET_MEAN, IMAGENET_STD, polygons2masks, polygons2masks_overlap
+
+
+# TODO: we might need a BaseTransform to make all these augments be compatible with both classification and semantic
+class BaseTransform:
+
+    def __init__(self) -> None:
+        pass
+
+    def apply_image(self, labels):
+        pass
+
+    def apply_instances(self, labels):
+        pass
+
+    def apply_semantic(self, labels):
+        pass
+
+    def __call__(self, labels):
+        self.apply_image(labels)
+        self.apply_instances(labels)
+        self.apply_semantic(labels)
+
+
+class Compose:
+
+    def __init__(self, transforms):
+        self.transforms = transforms
+
+    def __call__(self, data):
+        for t in self.transforms:
+            data = t(data)
+        return data
+
+    def append(self, transform):
+        self.transforms.append(transform)
+
+    def tolist(self):
+        return self.transforms
+
+    def __repr__(self):
+        format_string = f"{self.__class__.__name__}("
+        for t in self.transforms:
+            format_string += "\n"
+            format_string += f"    {t}"
+        format_string += "\n)"
+        return format_string
+
+
+class BaseMixTransform:
+    """This implementation is from mmyolo"""
+
+    def __init__(self, pre_transform=None, p=0.0) -> None:
+        self.pre_transform = pre_transform
+        self.p = p
+
+    def __call__(self, labels):
+        if random.uniform(0, 1) > self.p:
+            return labels
+
+        assert "dataset" in labels
+        dataset = labels.pop("dataset")
+
+        # get index of one or three other images
+        indexes = self.get_indexes(dataset)
+        if not isinstance(indexes, collections.abc.Sequence):
+            indexes = [indexes]
+
+        # get images information will be used for Mosaic or MixUp
+        mix_labels = [deepcopy(dataset.get_label_info(index)) for index in indexes]
+
+        if self.pre_transform is not None:
+            for i, data in enumerate(mix_labels):
+                # pre_transform may also require dataset
+                data.update({"dataset": dataset})
+                # before Mosaic or MixUp need to go through
+                # the necessary pre_transform
+                _labels = self.pre_transform(data)
+                _labels.pop("dataset")
+                mix_labels[i] = _labels
+        labels["mix_labels"] = mix_labels
+
+        # Mosaic or MixUp
+        labels = self._mix_transform(labels)
+
+        if "mix_labels" in labels:
+            labels.pop("mix_labels")
+        labels["dataset"] = dataset
+
+        return labels
+
+    def _mix_transform(self, labels):
+        raise NotImplementedError
+
+    def get_indexes(self, dataset):
+        raise NotImplementedError
+
+
+class Mosaic(BaseMixTransform):
+    """Mosaic augmentation.
+    Args:
+        img_size (Sequence[int]): Image size after mosaic pipeline of single
+            image. The shape order should be (height, width).
+            Default to (640, 640).
+    """
+
+    def __init__(self, img_size=640, p=1.0, border=(0, 0)):
+        assert 0 <= p <= 1.0, "The probability should be in range [0, 1]. " f"got {p}."
+        super().__init__(pre_transform=None, p=p)
+        self.img_size = img_size
+        self.border = border
+
+    def get_indexes(self, dataset):
+        return [random.randint(0, len(dataset)) for _ in range(3)]
+
+    def _mix_transform(self, labels):
+        mosaic_labels = []
+        assert labels.get("rect_shape", None) is None, "rect and mosaic is exclusive."
+        assert len(labels.get("mix_labels", [])) > 0, "There are no other images for mosaic augment."
+        s = self.img_size
+        yc, xc = (int(random.uniform(-x, 2 * s + x)) for x in self.border)  # mosaic center x, y
+        mix_labels = labels["mix_labels"]
+        for i in range(4):
+            labels_patch = deepcopy(labels) if i == 0 else deepcopy(mix_labels[i - 1])
+            # Load image
+            img = labels_patch["img"]
+            h, w = labels_patch["resized_shape"]
+
+            # place img in img4
+            if i == 0:  # top left
+                img4 = np.full((s * 2, s * 2, img.shape[2]), 114, dtype=np.uint8)  # base image with 4 tiles
+                x1a, y1a, x2a, y2a = max(xc - w, 0), max(yc - h, 0), xc, yc  # xmin, ymin, xmax, ymax (large image)
+                x1b, y1b, x2b, y2b = w - (x2a - x1a), h - (y2a - y1a), w, h  # xmin, ymin, xmax, ymax (small image)
+            elif i == 1:  # top right
+                x1a, y1a, x2a, y2a = xc, max(yc - h, 0), min(xc + w, s * 2), yc
+                x1b, y1b, x2b, y2b = 0, h - (y2a - y1a), min(w, x2a - x1a), h
+            elif i == 2:  # bottom left
+                x1a, y1a, x2a, y2a = max(xc - w, 0), yc, xc, min(s * 2, yc + h)
+                x1b, y1b, x2b, y2b = w - (x2a - x1a), 0, w, min(y2a - y1a, h)
+            elif i == 3:  # bottom right
+                x1a, y1a, x2a, y2a = xc, yc, min(xc + w, s * 2), min(s * 2, yc + h)
+                x1b, y1b, x2b, y2b = 0, 0, min(w, x2a - x1a), min(y2a - y1a, h)
+
+            img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b]  # img4[ymin:ymax, xmin:xmax]
+            padw = x1a - x1b
+            padh = y1a - y1b
+
+            labels_patch = self._update_labels(labels_patch, padw, padh)
+            mosaic_labels.append(labels_patch)
+        final_labels = self._cat_labels(mosaic_labels)
+        final_labels["img"] = img4
+        return final_labels
+
+    def _update_labels(self, labels, padw, padh):
+        """Update labels"""
+        nh, nw = labels["img"].shape[:2]
+        labels["instances"].convert_bbox(format="xyxy")
+        labels["instances"].denormalize(nw, nh)
+        labels["instances"].add_padding(padw, padh)
+        return labels
+
+    def _cat_labels(self, mosaic_labels):
+        if len(mosaic_labels) == 0:
+            return {}
+        cls = []
+        instances = []
+        for labels in mosaic_labels:
+            cls.append(labels["cls"])
+            instances.append(labels["instances"])
+        final_labels = {
+            "ori_shape": (self.img_size * 2, self.img_size * 2),
+            "resized_shape": (self.img_size * 2, self.img_size * 2),
+            "im_file": mosaic_labels[0]["im_file"],
+            "cls": np.concatenate(cls, 0)}
+
+        final_labels["instances"] = Instances.concatenate(instances, axis=0)
+        final_labels["instances"].clip(self.img_size * 2, self.img_size * 2)
+        return final_labels
+
+
+class MixUp(BaseMixTransform):
+
+    def __init__(self, pre_transform=None, p=0.0) -> None:
+        super().__init__(pre_transform=pre_transform, p=p)
+
+    def get_indexes(self, dataset):
+        return random.randint(0, len(dataset))
+
+    def _mix_transform(self, labels):
+        im = labels["img"]
+        labels2 = labels["mix_labels"][0]
+        im2 = labels2["img"]
+        cls2 = labels2["cls"]
+        # 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)
+        cat_instances = Instances.concatenate([labels["instances"], labels2["instances"]], axis=0)
+        cls = labels["cls"]
+        labels["img"] = im
+        labels["instances"] = cat_instances
+        labels["cls"] = np.concatenate([cls, cls2], 0)
+        return labels
+
+
+class RandomPerspective:
+
+    def __init__(self, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, border=(0, 0)):
+        self.degrees = degrees
+        self.translate = translate
+        self.scale = scale
+        self.shear = shear
+        self.perspective = perspective
+        # mosaic border
+        self.border = border
+
+    def affine_transform(self, img):
+        # 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(-self.perspective, self.perspective)  # x perspective (about y)
+        P[2, 1] = random.uniform(-self.perspective, self.perspective)  # y perspective (about x)
+
+        # Rotation and Scale
+        R = np.eye(3)
+        a = random.uniform(-self.degrees, self.degrees)
+        # a += random.choice([-180, -90, 0, 90])  # add 90deg rotations to small rotations
+        s = random.uniform(1 - self.scale, 1 + self.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(-self.shear, self.shear) * math.pi / 180)  # x shear (deg)
+        S[1, 0] = math.tan(random.uniform(-self.shear, self.shear) * math.pi / 180)  # y shear (deg)
+
+        # Translation
+        T = np.eye(3)
+        T[0, 2] = random.uniform(0.5 - self.translate, 0.5 + self.translate) * self.size[0]  # x translation (pixels)
+        T[1, 2] = random.uniform(0.5 - self.translate, 0.5 + self.translate) * self.size[1]  # y translation (pixels)
+
+        # Combined rotation matrix
+        M = T @ S @ R @ P @ C  # order of operations (right to left) is IMPORTANT
+        # affine image
+        if (self.border[0] != 0) or (self.border[1] != 0) or (M != np.eye(3)).any():  # image changed
+            if self.perspective:
+                img = cv2.warpPerspective(img, M, dsize=self.size, borderValue=(114, 114, 114))
+            else:  # affine
+                img = cv2.warpAffine(img, M[:2], dsize=self.size, borderValue=(114, 114, 114))
+        return img, M, s
+
+    def apply_bboxes(self, bboxes, M):
+        """apply affine to bboxes only.
+
+        Args:
+            bboxes(ndarray): list of bboxes, xyxy format, with shape (num_bboxes, 4).
+            M(ndarray): affine matrix.
+        Returns:
+            new_bboxes(ndarray): bboxes after affine, [num_bboxes, 4].
+        """
+        n = len(bboxes)
+        if n == 0:
+            return bboxes
+
+        xy = np.ones((n * 4, 3))
+        xy[:, :2] = bboxes[:, [0, 1, 2, 3, 0, 3, 2, 1]].reshape(n * 4, 2)  # x1y1, x2y2, x1y2, x2y1
+        xy = xy @ M.T  # transform
+        xy = (xy[:, :2] / xy[:, 2:3] if self.perspective else xy[:, :2]).reshape(n, 8)  # perspective rescale or affine
+
+        # create new boxes
+        x = xy[:, [0, 2, 4, 6]]
+        y = xy[:, [1, 3, 5, 7]]
+        return np.concatenate((x.min(1), y.min(1), x.max(1), y.max(1))).reshape(4, n).T
+
+    def apply_segments(self, segments, M):
+        """apply affine to segments and generate new bboxes from segments.
+
+        Args:
+            segments(ndarray): list of segments, [num_samples, 500, 2].
+            M(ndarray): affine matrix.
+        Returns:
+            new_segments(ndarray): list of segments after affine, [num_samples, 500, 2].
+            new_bboxes(ndarray): bboxes after affine, [N, 4].
+        """
+        n, num = segments.shape[:2]
+        if n == 0:
+            return [], segments
+
+        xy = np.ones((n * num, 3))
+        segments = segments.reshape(-1, 2)
+        xy[:, :2] = segments
+        xy = xy @ M.T  # transform
+        xy = xy[:, :2] / xy[:, 2:3]
+        segments = xy.reshape(n, -1, 2)
+        bboxes = np.stack([segment2box(xy, self.size[0], self.size[1]) for xy in segments], 0)
+        return bboxes, segments
+
+    def apply_keypoints(self, keypoints, M):
+        """apply affine to keypoints.
+
+        Args:
+            keypoints(ndarray): keypoints, [N, 17, 2].
+            M(ndarray): affine matrix.
+        Return:
+            new_keypoints(ndarray): keypoints after affine, [N, 17, 2].
+        """
+        n = len(keypoints)
+        if n == 0:
+            return keypoints
+        new_keypoints = np.ones((n * 17, 3))
+        new_keypoints[:, :2] = keypoints.reshape(n * 17, 2)  # num_kpt is hardcoded to 17
+        new_keypoints = new_keypoints @ M.T  # transform
+        new_keypoints = (new_keypoints[:, :2] / new_keypoints[:, 2:3]).reshape(n, 34)  # perspective rescale or affine
+        new_keypoints[keypoints.reshape(-1, 34) == 0] = 0
+        x_kpts = new_keypoints[:, list(range(0, 34, 2))]
+        y_kpts = new_keypoints[:, list(range(1, 34, 2))]
+
+        x_kpts[np.logical_or.reduce((x_kpts < 0, x_kpts > self.size[0], y_kpts < 0, y_kpts > self.size[1]))] = 0
+        y_kpts[np.logical_or.reduce((x_kpts < 0, x_kpts > self.size[0], y_kpts < 0, y_kpts > self.size[1]))] = 0
+        new_keypoints[:, list(range(0, 34, 2))] = x_kpts
+        new_keypoints[:, list(range(1, 34, 2))] = y_kpts
+        return new_keypoints.reshape(n, 17, 2)
+
+    def __call__(self, labels):
+        """
+        Affine images and targets.
+
+        Args:
+            img(ndarray): image.
+            labels(Dict): a dict of `bboxes`, `segments`, `keypoints`.
+        """
+        img = labels["img"]
+        cls = labels["cls"]
+        instances = labels["instances"]
+        # make sure the coord formats are right
+        instances.convert_bbox(format="xyxy")
+        instances.denormalize(*img.shape[:2][::-1])
+
+        self.size = img.shape[1] + self.border[1] * 2, img.shape[0] + self.border[0] * 2  # w, h
+        # M is affine matrix
+        # scale for func:`box_candidates`
+        img, M, scale = self.affine_transform(img)
+
+        bboxes = self.apply_bboxes(instances.bboxes, M)
+
+        segments = instances.segments
+        keypoints = instances.keypoints
+        # update bboxes if there are segments.
+        if segments is not None:
+            bboxes, segments = self.apply_segments(segments, M)
+
+        if keypoints is not None:
+            keypoints = self.apply_keypoints(keypoints, M)
+        new_instances = Instances(bboxes, segments, keypoints, bbox_format="xyxy", normalized=False)
+        new_instances.clip(*self.size)
+
+        # filter instances
+        instances.scale(scale_w=scale, scale_h=scale, bbox_only=True)
+        # make the bboxes have the same scale with new_bboxes
+        i = self.box_candidates(box1=instances.bboxes.T,
+                                box2=new_instances.bboxes.T,
+                                area_thr=0.01 if segments is not None else 0.10)
+        labels["instances"] = new_instances[i]
+        # clip
+        labels["cls"] = cls[i]
+        labels["img"] = img
+        return labels
+
+    def box_candidates(self, box1, box2, wh_thr=2, ar_thr=100, area_thr=0.1, eps=1e-16):  # 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 + eps), h2 / (w2 + eps))  # aspect ratio
+        return (w2 > wh_thr) & (h2 > wh_thr) & (w2 * h2 / (w1 * h1 + eps) > area_thr) & (ar < ar_thr)  # candidates
+
+
+class RandomHSV:
+
+    def __init__(self, hgain=0.5, sgain=0.5, vgain=0.5) -> None:
+        self.hgain = hgain
+        self.sgain = sgain
+        self.vgain = vgain
+
+    def __call__(self, labels):
+        img = labels["img"]
+        if self.hgain or self.sgain or self.vgain:
+            r = np.random.uniform(-1, 1, 3) * [self.hgain, self.sgain, self.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=r.dtype)
+            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)
+
+            im_hsv = cv2.merge((cv2.LUT(hue, lut_hue), cv2.LUT(sat, lut_sat), cv2.LUT(val, lut_val)))
+            cv2.cvtColor(im_hsv, cv2.COLOR_HSV2BGR, dst=img)  # no return needed
+        labels["img"] = img
+        return labels
+
+
+class RandomFlip:
+
+    def __init__(self, p=0.5, direction="horizontal") -> None:
+        assert direction in ["horizontal", "vertical"], f"Support direction `horizontal` or `vertical`, got {direction}"
+        assert 0 <= p <= 1.0
+
+        self.p = p
+        self.direction = direction
+
+    def __call__(self, labels):
+        img = labels["img"]
+        instances = labels["instances"]
+        instances.convert_bbox(format="xywh")
+        h, w = img.shape[:2]
+        h = 1 if instances.normalized else h
+        w = 1 if instances.normalized else w
+
+        # Flip up-down
+        if self.direction == "vertical" and random.random() < self.p:
+            img = np.flipud(img)
+            img = np.ascontiguousarray(img)
+            instances.flipud(h)
+        if self.direction == "horizontal" and random.random() < self.p:
+            img = np.fliplr(img)
+            img = np.ascontiguousarray(img)
+            instances.fliplr(w)
+        labels["img"] = img
+        labels["instances"] = instances
+        return labels
+
+
+class LetterBox:
+    """Resize image and padding for detection, instance segmentation, pose"""
+
+    def __init__(self, new_shape=(640, 640), auto=False, scaleFill=False, scaleup=True, stride=32):
+        self.new_shape = new_shape
+        self.auto = auto
+        self.scaleFill = scaleFill
+        self.scaleup = scaleup
+        self.stride = stride
+
+    def __call__(self, labels):
+        img = labels["img"]
+        shape = img.shape[:2]  # current shape [height, width]
+        new_shape = labels.get("rect_shape", self.new_shape)
+        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 self.scaleup:  # only scale down, do not scale up (for better val 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 self.auto:  # minimum rectangle
+            dw, dh = np.mod(dw, self.stride), np.mod(dh, self.stride)  # wh padding
+        elif self.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)
+        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=(114, 114, 114))  # add border
+
+        labels = self._update_labels(labels, ratio, dw, dh)
+        labels["img"] = img
+        return labels
+
+    def _update_labels(self, labels, ratio, padw, padh):
+        """Update labels"""
+        labels["instances"].convert_bbox(format="xyxy")
+        labels["instances"].denormalize(*labels["img"].shape[:2][::-1])
+        labels["instances"].scale(*ratio)
+        labels["instances"].add_padding(padw, padh)
+        return labels
+
+
+class CopyPaste:
+
+    def __init__(self, p=0.5) -> None:
+        self.p = p
+
+    def __call__(self, labels):
+        # Implement Copy-Paste augmentation https://arxiv.org/abs/2012.07177, labels as nx5 np.array(cls, xyxy)
+        im = labels["img"]
+        cls = labels["cls"]
+        bboxes = labels["instances"].bboxes
+        segments = labels["instances"].segments  # n, 1000, 2
+        keypoints = labels["instances"].keypoints
+        if self.p and segments is not None:
+            n = len(segments)
+            h, w, _ = im.shape  # height, width, channels
+            im_new = np.zeros(im.shape, np.uint8)
+            # TODO: this implement can be parallel since segments are ndarray, also might work with Instances inside
+            for j in random.sample(range(n), k=round(self.p * n)):
+                c, b, s = cls[j], bboxes[j], segments[j]
+                box = w - b[2], b[1], w - b[0], b[3]
+                ioa = bbox_ioa(box, bboxes)  # intersection over area
+                if (ioa < 0.30).all():  # allow 30% obscuration of existing labels
+                    bboxes = np.concatenate((bboxes, [box]), 0)
+                    cls = np.concatenate((cls, c[None]), axis=0)
+                    segments = np.concatenate((segments, np.concatenate((w - s[:, 0:1], s[:, 1:2]), 1)[None]), 0)
+                    if keypoints is not None:
+                        keypoints = np.concatenate(
+                            (keypoints, np.concatenate((w - keypoints[j][:, 0:1], keypoints[j][:, 1:2]), 1)), 0)
+                    cv2.drawContours(im_new, [segments[j].astype(np.int32)], -1, (255, 255, 255), cv2.FILLED)
+
+            result = cv2.bitwise_and(src1=im, src2=im_new)
+            result = cv2.flip(result, 1)  # augment segments (flip left-right)
+            i = result > 0  # pixels to replace
+            # i[:, :] = result.max(2).reshape(h, w, 1)  # act over ch
+            im[i] = result[i]  # cv2.imwrite('debug.jpg', im)  # debug
+        labels["img"] = im
+        labels["cls"] = cls
+        labels["instances"].update(bboxes, segments, keypoints)
+        return labels
+
+
+class Albumentations:
+    # YOLOv5 Albumentations class (optional, only used if package is installed)
+    def __init__(self, p=1.0):
+        self.p = p
+        self.transform = None
+        prefix = colorstr("albumentations: ")
+        try:
+            import albumentations as A
+
+            check_version(A.__version__, "1.0.3", hard=True)  # version requirement
+
+            T = [
+                A.Blur(p=0.01),
+                A.MedianBlur(p=0.01),
+                A.ToGray(p=0.01),
+                A.CLAHE(p=0.01),
+                A.RandomBrightnessContrast(p=0.0),
+                A.RandomGamma(p=0.0),
+                A.ImageCompression(quality_lower=75, p=0.0),]  # transforms
+            self.transform = A.Compose(T, bbox_params=A.BboxParams(format="yolo", label_fields=["class_labels"]))
+
+            LOGGER.info(prefix + ", ".join(f"{x}".replace("always_apply=False, ", "") for x in T if x.p))
+        except ImportError:  # package not installed, skip
+            pass
+        except Exception as e:
+            LOGGER.info(f"{prefix}{e}")
+
+    def __call__(self, labels):
+        im = labels["img"]
+        cls = labels["cls"]
+        if len(cls):
+            labels["instances"].convert_bbox("xywh")
+            labels["instances"].normalize(*im.shape[:2][::-1])
+            bboxes = labels["instances"].bboxes
+            # TODO: add supports of segments and keypoints
+            if self.transform and random.random() < self.p:
+                new = self.transform(image=im, bboxes=bboxes, class_labels=cls)  # transformed
+            labels["img"] = new["image"]
+            labels["cls"] = np.array(new["class_labels"])
+            labels["instances"].update(bboxes=bboxes)
+        return labels
+
+
+# TODO: technically this is not an augmentation, maybe we should put this to another files
+class Format:
+
+    def __init__(self, bbox_format="xywh", normalize=True, mask=False, mask_ratio=4, mask_overlap=True, batch_idx=True):
+        self.bbox_format = bbox_format
+        self.normalize = normalize
+        self.mask = mask  # set False when training detection only
+        self.mask_ratio = mask_ratio
+        self.mask_overlap = mask_overlap
+        self.batch_idx = batch_idx  # keep the batch indexes
+
+    def __call__(self, labels):
+        img = labels["img"]
+        h, w = img.shape[:2]
+        cls = labels.pop("cls")
+        instances = labels.pop("instances")
+        instances.convert_bbox(format=self.bbox_format)
+        instances.denormalize(w, h)
+        nl = len(instances)
+
+        if instances.segments is not None and self.mask:
+            masks, instances, cls = self._format_segments(instances, cls, w, h)
+            labels["masks"] = (torch.from_numpy(masks) if nl else torch.zeros(
+                1 if self.mask_overlap else nl, img.shape[0] // self.mask_ratio, img.shape[1] // self.mask_ratio))
+        if self.normalize:
+            instances.normalize(w, h)
+        labels["img"] = self._format_img(img)
+        labels["cls"] = torch.from_numpy(cls) if nl else torch.zeros(nl)
+        labels["bboxes"] = torch.from_numpy(instances.bboxes) if nl else torch.zeros((nl, 4))
+        if instances.keypoints is not None:
+            labels["keypoints"] = torch.from_numpy(instances.keypoints) if nl else torch.zeros((nl, 17, 2))
+        # then we can use collate_fn
+        if self.batch_idx:
+            labels["batch_idx"] = torch.zeros(nl)
+        return labels
+
+    def _format_img(self, img):
+        if len(img.shape) < 3:
+            img = np.expand_dims(img, -1)
+        img = np.ascontiguousarray(img.transpose(2, 0, 1))
+        img = torch.from_numpy(img)
+        return img
+
+    def _format_segments(self, instances, cls, w, h):
+        """convert polygon points to bitmap"""
+        segments = instances.segments
+        if self.mask_overlap:
+            masks, sorted_idx = polygons2masks_overlap((h, w), segments, downsample_ratio=self.mask_ratio)
+            masks = masks[None]  # (640, 640) -> (1, 640, 640)
+            instances = instances[sorted_idx]
+            cls = cls[sorted_idx]
+        else:
+            masks = polygons2masks((h, w), segments, color=1, downsample_ratio=self.mask_ratio)
+
+        return masks, instances, cls
+
+
+def mosaic_transforms(img_size, hyp):
+    pre_transform = Compose([
+        Mosaic(img_size=img_size, p=hyp.mosaic, border=[-img_size // 2, -img_size // 2]),
+        CopyPaste(p=hyp.copy_paste),
+        RandomPerspective(
+            degrees=hyp.degrees,
+            translate=hyp.translate,
+            scale=hyp.scale,
+            shear=hyp.shear,
+            perspective=hyp.perspective,
+            border=[-img_size // 2, -img_size // 2],
+        ),])
+    transforms = Compose([
+        pre_transform,
+        MixUp(
+            pre_transform=pre_transform,
+            p=hyp.mixup,
+        ),
+        Albumentations(p=1.0),
+        RandomHSV(hgain=hyp.hsv_h, sgain=hyp.hsv_s, vgain=hyp.hsv_v),
+        RandomFlip(direction="vertical", p=hyp.flipud),
+        RandomFlip(direction="horizontal", p=hyp.fliplr),])
+    return transforms
+
+
+def affine_transforms(img_size, hyp):
+    # rect, randomperspective, albumentation, hsv, flipud, fliplr
+    transforms = Compose([
+        LetterBox(new_shape=(img_size, img_size)),
+        RandomPerspective(
+            degrees=hyp.degrees,
+            translate=hyp.translate,
+            scale=hyp.scale,
+            shear=hyp.shear,
+            perspective=hyp.perspective,
+            border=[0, 0],
+        ),
+        Albumentations(p=1.0),
+        RandomHSV(hgain=hyp.hsv_h, sgain=hyp.hsv_s, vgain=hyp.hsv_v),
+        RandomFlip(direction="vertical", p=hyp.flipud),
+        RandomFlip(direction="horizontal", p=hyp.fliplr),])
+    return transforms
+
+
+# Classification augmentations -------------------------------------------------------------------------------------------
+def classify_transforms(size=224):
+    # Transforms to apply if albumentations not installed
+    assert isinstance(size, int), f"ERROR: classify_transforms size {size} must be integer, not (list, tuple)"
+    # T.Compose([T.ToTensor(), T.Resize(size), T.CenterCrop(size), T.Normalize(IMAGENET_MEAN, IMAGENET_STD)])
+    return T.Compose([CenterCrop(size), ToTensor(), T.Normalize(IMAGENET_MEAN, IMAGENET_STD)])
+
+
+def classify_albumentations(
+        augment=True,
+        size=224,
+        scale=(0.08, 1.0),
+        hflip=0.5,
+        vflip=0.0,
+        jitter=0.4,
+        mean=IMAGENET_MEAN,
+        std=IMAGENET_STD,
+        auto_aug=False,
+):
+    # YOLOv5 classification Albumentations (optional, only used if package is installed)
+    prefix = colorstr("albumentations: ")
+    try:
+        import albumentations as A
+        from albumentations.pytorch import ToTensorV2
+
+        check_version(A.__version__, "1.0.3", hard=True)  # version requirement
+        if augment:  # Resize and crop
+            T = [A.RandomResizedCrop(height=size, width=size, scale=scale)]
+            if auto_aug:
+                # TODO: implement AugMix, AutoAug & RandAug in albumentation
+                LOGGER.info(f"{prefix}auto augmentations are currently not supported")
+            else:
+                if hflip > 0:
+                    T += [A.HorizontalFlip(p=hflip)]
+                if vflip > 0:
+                    T += [A.VerticalFlip(p=vflip)]
+                if jitter > 0:
+                    color_jitter = (float(jitter),) * 3  # repeat value for brightness, contrast, satuaration, 0 hue
+                    T += [A.ColorJitter(*color_jitter, 0)]
+        else:  # Use fixed crop for eval set (reproducibility)
+            T = [A.SmallestMaxSize(max_size=size), A.CenterCrop(height=size, width=size)]
+        T += [A.Normalize(mean=mean, std=std), ToTensorV2()]  # Normalize and convert to Tensor
+        LOGGER.info(prefix + ", ".join(f"{x}".replace("always_apply=False, ", "") for x in T if x.p))
+        return A.Compose(T)
+
+    except ImportError:  # package not installed, skip
+        pass
+    except Exception as e:
+        LOGGER.info(f"{prefix}{e}")
+
+
+class ClassifyLetterBox:
+    # YOLOv5 LetterBox class for image preprocessing, i.e. T.Compose([LetterBox(size), ToTensor()])
+    def __init__(self, size=(640, 640), auto=False, stride=32):
+        super().__init__()
+        self.h, self.w = (size, size) if isinstance(size, int) else size
+        self.auto = auto  # pass max size integer, automatically solve for short side using stride
+        self.stride = stride  # used with auto
+
+    def __call__(self, im):  # im = np.array HWC
+        imh, imw = im.shape[:2]
+        r = min(self.h / imh, self.w / imw)  # ratio of new/old
+        h, w = round(imh * r), round(imw * r)  # resized image
+        hs, ws = (math.ceil(x / self.stride) * self.stride for x in (h, w)) if self.auto else self.h, self.w
+        top, left = round((hs - h) / 2 - 0.1), round((ws - w) / 2 - 0.1)
+        im_out = np.full((self.h, self.w, 3), 114, dtype=im.dtype)
+        im_out[top:top + h, left:left + w] = cv2.resize(im, (w, h), interpolation=cv2.INTER_LINEAR)
+        return im_out
+
+
+class CenterCrop:
+    # YOLOv5 CenterCrop class for image preprocessing, i.e. T.Compose([CenterCrop(size), ToTensor()])
+    def __init__(self, size=640):
+        super().__init__()
+        self.h, self.w = (size, size) if isinstance(size, int) else size
+
+    def __call__(self, im):  # im = np.array HWC
+        imh, imw = im.shape[:2]
+        m = min(imh, imw)  # min dimension
+        top, left = (imh - m) // 2, (imw - m) // 2
+        return cv2.resize(im[top:top + m, left:left + m], (self.w, self.h), interpolation=cv2.INTER_LINEAR)
+
+
+class ToTensor:
+    # YOLOv5 ToTensor class for image preprocessing, i.e. T.Compose([LetterBox(size), ToTensor()])
+    def __init__(self, half=False):
+        super().__init__()
+        self.half = half
+
+    def __call__(self, im):  # im = np.array HWC in BGR order
+        im = np.ascontiguousarray(im.transpose((2, 0, 1))[::-1])  # HWC to CHW -> BGR to RGB -> contiguous
+        im = torch.from_numpy(im)  # to torch
+        im = im.half() if self.half else im.float()  # uint8 to fp16/32
+        im /= 255.0  # 0-255 to 0.0-1.0
+        return im

ultralytics/yolo/data/base.py

@@ -0,0 +1,224 @@
+import glob
+import os
+from multiprocessing.pool import ThreadPool
+from pathlib import Path
+from typing import Optional
+
+import cv2
+import numpy as np
+from torch.utils.data import Dataset
+from tqdm import tqdm
+
+from ..utils.general import NUM_THREADS
+from .utils import BAR_FORMAT, HELP_URL, IMG_FORMATS, LOCAL_RANK
+
+
+class BaseDataset(Dataset):
+    """Base Dataset.
+    Args:
+        img_path (str): image path.
+        pipeline (dict): a dict of image transforms.
+        label_path (str): label path, this can also be a ann_file or other custom label path.
+    """
+
+    def __init__(
+        self,
+        img_path,
+        img_size=640,
+        label_path=None,
+        cache=False,
+        augment=True,
+        hyp=None,
+        prefix="",
+        rect=False,
+        batch_size=None,
+        stride=32,
+        pad=0.5,
+        single_cls=False,
+    ):
+        super().__init__()
+        self.img_path = img_path
+        self.img_size = img_size
+        self.label_path = label_path
+        self.augment = augment
+        self.prefix = prefix
+
+        self.im_files = self.get_img_files(self.img_path)
+        self.labels = self.get_labels()
+        if single_cls:
+            self.update_labels(include_class=[], single_cls=single_cls)
+
+        self.ni = len(self.im_files)
+
+        # rect stuff
+        self.rect = rect
+        self.batch_size = batch_size
+        self.stride = stride
+        self.pad = pad
+        if self.rect:
+            assert self.batch_size is not None
+            self.set_rectangle()
+
+        # cache stuff
+        self.ims = [None] * self.ni
+        self.npy_files = [Path(f).with_suffix(".npy") for f in self.im_files]
+        if cache:
+            self.cache_images()
+
+        # transforms
+        self.transforms = self.build_transforms(hyp=hyp)
+
+    def get_img_files(self, img_path):
+        """Read image files."""
+        try:
+            f = []  # image files
+            for p in img_path if isinstance(img_path, list) else [img_path]:
+                p = Path(p)  # os-agnostic
+                if p.is_dir():  # dir
+                    f += glob.glob(str(p / "**" / "*.*"), recursive=True)
+                    # f = list(p.rglob('*.*'))  # pathlib
+                elif p.is_file():  # file
+                    with open(p) as t:
+                        t = t.read().strip().splitlines()
+                        parent = str(p.parent) + os.sep
+                        f += [x.replace("./", parent) if x.startswith("./") else x for x in t]  # local to global path
+                        # f += [p.parent / x.lstrip(os.sep) for x in t]  # local to global path (pathlib)
+                else:
+                    raise FileNotFoundError(f"{self.prefix}{p} does not exist")
+            im_files = sorted(x.replace("/", os.sep) for x in f if x.split(".")[-1].lower() in IMG_FORMATS)
+            # self.img_files = sorted([x for x in f if x.suffix[1:].lower() in IMG_FORMATS])  # pathlib
+            assert im_files, f"{self.prefix}No images found"
+        except Exception as e:
+            raise Exception(f"{self.prefix}Error loading data from {img_path}: {e}\n{HELP_URL}")
+        return im_files
+
+    def update_labels(self, include_class: Optional[list]):
+        """include_class, filter labels to include only these classes (optional)"""
+        include_class_array = np.array(include_class).reshape(1, -1)
+        for i in range(len(self.labels)):
+            if include_class:
+                cls = self.labels[i]["cls"]
+                bboxes = self.labels[i]["bboxes"]
+                segments = self.labels[i]["segments"]
+                j = (cls == include_class_array).any(1)
+                self.labels[i]["cls"] = cls[j]
+                self.labels[i]["bboxes"] = bboxes[j]
+                if segments:
+                    self.labels[i]["segments"] = segments[j]
+            if self.single_cls:
+                self.labels[i]["cls"] = 0
+
+    def load_image(self, i):
+        # Loads 1 image from dataset index 'i', returns (im, resized hw)
+        im, f, fn = self.ims[i], self.im_files[i], self.npy_files[i]
+        if im is None:  # not cached in RAM
+            if fn.exists():  # load npy
+                im = np.load(fn)
+            else:  # read image
+                im = cv2.imread(f)  # BGR
+                assert im is not None, f"Image Not Found {f}"
+            h0, w0 = im.shape[:2]  # orig hw
+            r = self.img_size / max(h0, w0)  # ratio
+            if r != 1:  # if sizes are not equal
+                interp = cv2.INTER_LINEAR if (self.augment or r > 1) else cv2.INTER_AREA
+                im = cv2.resize(im, (int(w0 * r), int(h0 * r)), interpolation=interp)
+            return im, (h0, w0), im.shape[:2]  # im, hw_original, hw_resized
+        return self.ims[i], self.im_hw0[i], self.im_hw[i]  # im, hw_original, hw_resized
+
+    def cache_images(self):
+        # cache images to memory or disk
+        gb = 0  # Gigabytes of cached images
+        self.im_hw0, self.im_hw = [None] * self.ni, [None] * self.ni
+        fcn = self.cache_images_to_disk if self.cache == "disk" else self.load_image
+        results = ThreadPool(NUM_THREADS).imap(fcn, range(self.ni))
+        pbar = tqdm(enumerate(results), total=self.ni, bar_format=BAR_FORMAT, disable=LOCAL_RANK > 0)
+        for i, x in pbar:
+            if self.cache == "disk":
+                gb += self.npy_files[i].stat().st_size
+            else:  # 'ram'
+                self.ims[i], self.im_hw0[i], self.im_hw[i] = x  # im, hw_orig, hw_resized = load_image(self, i)
+                gb += self.ims[i].nbytes
+            pbar.desc = f"{self.prefix}Caching images ({gb / 1E9:.1f}GB {self.cache})"
+        pbar.close()
+
+    def cache_images_to_disk(self, i):
+        # Saves an image as an *.npy file for faster loading
+        f = self.npy_files[i]
+        if not f.exists():
+            np.save(f.as_posix(), cv2.imread(self.im_files[i]))
+
+    def set_rectangle(self):
+        bi = np.floor(np.arange(self.ni) / self.batch_size).astype(int)  # batch index
+        nb = bi[-1] + 1  # number of batches
+
+        s = np.array([x["shape"] for x in self.labels])  # hw
+        ar = s[:, 0] / s[:, 1]  # aspect ratio
+        irect = ar.argsort()
+        self.im_files = [self.im_files[i] for i in irect]
+        self.labels = [self.labels[i] for i in irect]
+        ar = ar[irect]
+
+        # Set training image shapes
+        shapes = [[1, 1]] * nb
+        for i in range(nb):
+            ari = ar[bi == i]
+            mini, maxi = ari.min(), ari.max()
+            if maxi < 1:
+                shapes[i] = [maxi, 1]
+            elif mini > 1:
+                shapes[i] = [1, 1 / mini]
+
+        self.batch_shapes = np.ceil(np.array(shapes) * self.img_size / self.stride + self.pad).astype(int) * self.stride
+        self.batch = bi  # batch index of image
+
+    def __getitem__(self, index):
+        label = self.get_label_info(index)
+        if self.augment:
+            label["dataset"] = self
+        return self.transforms(label)
+
+    def get_label_info(self, index):
+        label = self.labels[index].copy()
+        img, (h0, w0), (h, w) = self.load_image(index)
+        label["img"] = img
+        label["ori_shape"] = (h0, w0)
+        label["resized_shape"] = (h, w)
+        if self.rect:
+            label["rect_shape"] = self.batch_shapes[self.batch[index]]
+        label = self.update_labels_info(label)
+        return label
+
+    def __len__(self):
+        return len(self.im_files)
+
+    def update_labels_info(self, label):
+        """custom your label format here"""
+        return label
+
+    def build_transforms(self, hyp=None):
+        """Users can custom augmentations here
+        like:
+            if self.augment:
+                # training transforms
+                return Compose([])
+            else:
+                # val transforms
+                return Compose([])
+        """
+        raise NotImplementedError
+
+    def get_labels(self):
+        """Users can custom their own format here.
+        Make sure your output is a list with each element like below:
+            dict(
+                im_file=im_file,
+                shape=shape,  # format: (height, width)
+                cls=cls,
+                bboxes=bboxes, # xywh
+                segments=segments,  # xy
+                keypoints=keypoints, # xy
+                normalized=True, # or False
+                bbox_format="xyxy",  # or xywh, ltwh
+            )
+        """
+        raise NotImplementedError

ultralytics/yolo/data/build.py

@@ -0,0 +1,145 @@
+import os
+import random
+
+import numpy as np
+import torch
+from torch.utils.data import DataLoader, dataloader, distributed
+
+from ..utils.general import LOGGER
+from ..utils.torch_utils import torch_distributed_zero_first
+from .dataset import ClassificationDataset, YOLODataset
+from .utils import PIN_MEMORY, RANK
+
+
+class InfiniteDataLoader(dataloader.DataLoader):
+    """Dataloader that reuses workers
+
+    Uses same syntax as vanilla DataLoader
+    """
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        object.__setattr__(self, "batch_sampler", _RepeatSampler(self.batch_sampler))
+        self.iterator = super().__iter__()
+
+    def __len__(self):
+        return len(self.batch_sampler.sampler)
+
+    def __iter__(self):
+        for _ in range(len(self)):
+            yield next(self.iterator)
+
+
+class _RepeatSampler:
+    """Sampler that repeats forever
+
+    Args:
+        sampler (Sampler)
+    """
+
+    def __init__(self, sampler):
+        self.sampler = sampler
+
+    def __iter__(self):
+        while True:
+            yield from iter(self.sampler)
+
+
+def seed_worker(worker_id):
+    # Set dataloader worker seed https://pytorch.org/docs/stable/notes/randomness.html#dataloader
+    worker_seed = torch.initial_seed() % 2 ** 32
+    np.random.seed(worker_seed)
+    random.seed(worker_seed)
+
+
+# TODO: we can inject most args from a config file
+def build_dataloader(
+    img_path,
+    img_size,  #
+    batch_size,  #
+    single_cls=False,  #
+    hyp=None,  #
+    augment=False,
+    cache=False,  #
+    image_weights=False,  #
+    stride=32,
+    label_path=None,
+    pad=0.0,
+    rect=False,
+    rank=-1,
+    workers=8,
+    prefix="",
+    shuffle=False,
+    use_segments=False,
+    use_keypoints=False,
+):
+    if rect and shuffle:
+        LOGGER.warning("WARNING ⚠️ --rect is incompatible with DataLoader shuffle, setting shuffle=False")
+        shuffle = False
+    with torch_distributed_zero_first(rank):  # init dataset *.cache only once if DDP
+        dataset = YOLODataset(
+            img_path=img_path,
+            img_size=img_size,
+            batch_size=batch_size,
+            label_path=label_path,
+            augment=augment,  # augmentation
+            hyp=hyp,
+            rect=rect,  # rectangular batches
+            cache=cache,
+            single_cls=single_cls,
+            stride=int(stride),
+            pad=pad,
+            prefix=prefix,
+            use_segments=use_segments,
+            use_keypoints=use_keypoints,
+        )
+
+    batch_size = min(batch_size, len(dataset))
+    nd = torch.cuda.device_count()  # number of CUDA devices
+    nw = min([os.cpu_count() // max(nd, 1), batch_size if batch_size > 1 else 0, workers])  # number of workers
+    sampler = None if rank == -1 else distributed.DistributedSampler(dataset, shuffle=shuffle)
+    loader = DataLoader if image_weights else InfiniteDataLoader  # only DataLoader allows for attribute updates
+    generator = torch.Generator()
+    generator.manual_seed(6148914691236517205 + RANK)
+    return (
+        loader(
+            dataset=dataset,
+            batch_size=batch_size,
+            shuffle=shuffle and sampler is None,
+            num_workers=nw,
+            sampler=sampler,
+            pin_memory=PIN_MEMORY,
+            collate_fn=getattr(dataset, "collate_fn", None),
+            worker_init_fn=seed_worker,
+            generator=generator,
+        ),
+        dataset,
+    )
+
+
+# build classification
+def build_classification_dataloader(path,
+                                    imgsz=224,
+                                    batch_size=16,
+                                    augment=True,
+                                    cache=False,
+                                    rank=-1,
+                                    workers=8,
+                                    shuffle=True):
+    # Returns Dataloader object to be used with YOLOv5 Classifier
+    with torch_distributed_zero_first(rank):  # init dataset *.cache only once if DDP
+        dataset = ClassificationDataset(root=path, imgsz=imgsz, augment=augment, cache=cache)
+    batch_size = min(batch_size, len(dataset))
+    nd = torch.cuda.device_count()
+    nw = min([os.cpu_count() // max(nd, 1), batch_size if batch_size > 1 else 0, workers])
+    sampler = None if rank == -1 else distributed.DistributedSampler(dataset, shuffle=shuffle)
+    generator = torch.Generator()
+    generator.manual_seed(6148914691236517205 + RANK)
+    return InfiniteDataLoader(dataset,
+                              batch_size=batch_size,
+                              shuffle=shuffle and sampler is None,
+                              num_workers=nw,
+                              sampler=sampler,
+                              pin_memory=PIN_MEMORY,
+                              worker_init_fn=seed_worker,
+                              generator=generator)  # or DataLoader(persistent_workers=True)

ultralytics/yolo/data/dataset.py

@@ -0,0 +1,213 @@
+from itertools import repeat
+from multiprocessing.pool import Pool
+from pathlib import Path
+
+import cv2
+import numpy as np
+import torch
+import torchvision
+from tqdm import tqdm
+
+from ..utils.general import LOGGER, NUM_THREADS
+from .augment import *
+from .base import BaseDataset
+from .utils import BAR_FORMAT, HELP_URL, LOCAL_RANK, get_hash, img2label_paths, verify_image_label
+
+
+class YOLODataset(BaseDataset):
+    cache_version = 0.6  # dataset labels *.cache version
+    rand_interp_methods = [cv2.INTER_NEAREST, cv2.INTER_LINEAR, cv2.INTER_CUBIC, cv2.INTER_AREA, cv2.INTER_LANCZOS4]
+    """YOLO Dataset.
+    Args:
+        img_path (str): image path.
+        prefix (str): prefix.
+    """
+
+    def __init__(
+        self,
+        img_path,
+        img_size=640,
+        label_path=None,
+        cache=False,
+        augment=True,
+        hyp=None,
+        prefix="",
+        rect=False,
+        batch_size=None,
+        stride=32,
+        pad=0.0,
+        single_cls=False,
+        use_segments=False,
+        use_keypoints=False,
+    ):
+        self.use_segments = use_segments
+        self.use_keypoints = use_keypoints
+        assert not (self.use_segments and self.use_keypoints), "We can't use both of segmentation and pose."
+        super().__init__(img_path, img_size, label_path, cache, augment, hyp, prefix, rect, batch_size, stride, pad,
+                         single_cls)
+
+    def cache_labels(self, path=Path("./labels.cache")):
+        # Cache dataset labels, check images and read shapes
+        x = {"labels": []}
+        nm, nf, ne, nc, msgs = 0, 0, 0, 0, []  # number missing, found, empty, corrupt, messages
+        desc = f"{self.prefix}Scanning '{path.parent / path.stem}' images and labels..."
+        with Pool(NUM_THREADS) as pool:
+            pbar = tqdm(
+                pool.imap(verify_image_label,
+                          zip(self.im_files, self.label_files, repeat(self.prefix), repeat(self.use_keypoints))),
+                desc=desc,
+                total=len(self.im_files),
+                bar_format=BAR_FORMAT,
+            )
+            for im_file, lb, shape, segments, keypoint, nm_f, nf_f, ne_f, nc_f, msg in pbar:
+                nm += nm_f
+                nf += nf_f
+                ne += ne_f
+                nc += nc_f
+                if im_file:
+                    x["labels"].append(
+                        dict(
+                            im_file=im_file,
+                            shape=shape,
+                            cls=lb[:, 0:1],  # n, 1
+                            bboxes=lb[:, 1:],  # n, 4
+                            segments=segments,
+                            keypoints=keypoint,
+                            normalized=True,
+                            bbox_format="xywh",
+                        ))
+                if msg:
+                    msgs.append(msg)
+                pbar.desc = f"{desc}{nf} found, {nm} missing, {ne} empty, {nc} corrupt"
+
+        pbar.close()
+        if msgs:
+            LOGGER.info("\n".join(msgs))
+        if nf == 0:
+            LOGGER.warning(f"{self.prefix}WARNING ⚠️ No labels found in {path}. {HELP_URL}")
+        x["hash"] = get_hash(self.label_files + self.im_files)
+        x["results"] = nf, nm, ne, nc, len(self.im_files)
+        x["msgs"] = msgs  # warnings
+        x["version"] = self.cache_version  # cache version
+        try:
+            np.save(path, x)  # save cache for next time
+            path.with_suffix(".cache.npy").rename(path)  # remove .npy suffix
+            LOGGER.info(f"{self.prefix}New cache created: {path}")
+        except Exception as e:
+            LOGGER.warning(
+                f"{self.prefix}WARNING ⚠️ Cache directory {path.parent} is not writeable: {e}")  # not writeable
+        return x
+
+    def get_labels(self):
+        self.label_files = img2label_paths(self.im_files)
+        cache_path = Path(self.label_files[0]).parent.with_suffix(".cache")
+        try:
+            cache, exists = np.load(cache_path, allow_pickle=True).item(), True  # load dict
+            assert cache["version"] == self.cache_version  # matches current version
+            assert cache["hash"] == get_hash(self.label_files + self.im_files)  # identical hash
+        except Exception:
+            cache, exists = self.cache_labels(cache_path), False  # run cache ops
+
+        # Display cache
+        nf, nm, ne, nc, n = cache.pop("results")  # found, missing, empty, corrupt, total
+        if exists and LOCAL_RANK in {-1, 0}:
+            d = f"Scanning '{cache_path}' images and labels... {nf} found, {nm} missing, {ne} empty, {nc} corrupt"
+            tqdm(None, desc=self.prefix + d, total=n, initial=n, bar_format=BAR_FORMAT)  # display cache results
+            if cache["msgs"]:
+                LOGGER.info("\n".join(cache["msgs"]))  # display warnings
+        assert nf > 0, f"{self.prefix}No labels found in {cache_path}, can not start training. {HELP_URL}"
+
+        # Read cache
+        [cache.pop(k) for k in ("hash", "version", "msgs")]  # remove items
+        labels = cache["labels"]
+        nl = len(np.concatenate([label["cls"] for label in labels], 0))  # number of labels
+        assert nl > 0, f"{self.prefix}All labels empty in {cache_path}, can not start training. {HELP_URL}"
+        return labels
+
+    # TODO: use hyp config to set all these augmentations
+    def build_transforms(self, hyp=None):
+        mosaic = self.augment and not self.rect
+        # mosaic = False
+        if self.augment:
+            if mosaic:
+                transforms = mosaic_transforms(self.img_size, hyp)
+            else:
+                transforms = affine_transforms(self.img_size, hyp)
+        else:
+            transforms = Compose([LetterBox(new_shape=(self.img_size, self.img_size))])
+        transforms.append(Format(bbox_format="xywh", normalize=True, mask=self.use_segments, batch_idx=True))
+        return transforms
+
+    def update_labels_info(self, label):
+        """custom your label format here"""
+        # NOTE: cls is not with bboxes now, since other tasks like classification and semantic segmentation need a independent cls label
+        # we can make it also support classification and semantic segmentation by add or remove some dict keys there.
+        bboxes = label.pop("bboxes")
+        segments = label.pop("segments", None)
+        keypoints = label.pop("keypoints", None)
+        bbox_format = label.pop("bbox_format")
+        normalized = label.pop("normalized")
+        label["instances"] = Instances(bboxes, segments, keypoints, bbox_format=bbox_format, normalized=normalized)
+        return label
+
+    @staticmethod
+    def collate_fn(batch):
+        # TODO: returning a dict can make thing easier and cleaner when using dataset in training
+        # but I don't know if this will slow down a little bit.
+        new_batch = {}
+        keys = batch[0].keys()
+        values = list(zip(*[list(b.values()) for b in batch]))
+        for i, k in enumerate(keys):
+            value = values[i]
+            if k == "img":
+                value = torch.stack(value, 0)
+            if k in ["mask", "keypoint", "bboxes", "cls"]:
+                value = torch.cat(value, 0)
+            new_batch[k] = values[i]
+        new_batch["batch_idx"] = list(new_batch["batch_idx"])
+        for i in range(len(new_batch["batch_idx"])):
+            new_batch["batch_idx"][i] += i  # add target image index for build_targets()
+        new_batch["batch_idx"] = torch.cat(new_batch["batch_idx"], 0)
+        return new_batch
+
+
+# Classification dataloaders -------------------------------------------------------------------------------------------
+class ClassificationDataset(torchvision.datasets.ImageFolder):
+    """
+    YOLOv5 Classification Dataset.
+    Arguments
+        root:  Dataset path
+        transform:  torchvision transforms, used by default
+        album_transform: Albumentations transforms, used if installed
+    """
+
+    def __init__(self, root, augment, imgsz, cache=False):
+        super().__init__(root=root)
+        self.torch_transforms = classify_transforms(imgsz)
+        self.album_transforms = classify_albumentations(augment, imgsz) if augment else None
+        self.cache_ram = cache is True or cache == "ram"
+        self.cache_disk = cache == "disk"
+        self.samples = [list(x) + [Path(x[0]).with_suffix(".npy"), None] for x in self.samples]  # file, index, npy, im
+
+    def __getitem__(self, i):
+        f, j, fn, im = self.samples[i]  # filename, index, filename.with_suffix('.npy'), image
+        if self.cache_ram and im is None:
+            im = self.samples[i][3] = cv2.imread(f)
+        elif self.cache_disk:
+            if not fn.exists():  # load npy
+                np.save(fn.as_posix(), cv2.imread(f))
+            im = np.load(fn)
+        else:  # read image
+            im = cv2.imread(f)  # BGR
+        if self.album_transforms:
+            sample = self.album_transforms(image=cv2.cvtColor(im, cv2.COLOR_BGR2RGB))["image"]
+        else:
+            sample = self.torch_transforms(im)
+        return sample, j
+
+
+# TODO: support semantic segmentation
+class SemanticDataset(BaseDataset):
+
+    def __init__(self):
+        pass

ultralytics/yolo/data/dataset_wrappers.py

@@ -0,0 +1,37 @@
+import collections
+from copy import deepcopy
+
+from .augment import LetterBox
+
+
+class MixAndRectDataset:
+    """A wrapper of multiple images mixed dataset.
+
+    Args:
+        dataset (:obj:`BaseDataset`): The dataset to be mixed.
+        transforms (Sequence[dict]): config dict to be composed.
+    """
+
+    def __init__(self, dataset):
+        self.dataset = dataset
+        self.img_size = dataset.img_size
+
+    def __len__(self):
+        return len(self.dataset)
+
+    def __getitem__(self, index):
+        labels = deepcopy(self.dataset[index])
+        for transform in self.dataset.transforms.tolist():
+            # mosaic and mixup
+            if hasattr(transform, "get_indexes"):
+                indexes = transform.get_indexes(self.dataset)
+                if not isinstance(indexes, collections.abc.Sequence):
+                    indexes = [indexes]
+                mix_labels = [deepcopy(self.dataset[index]) for index in indexes]
+                labels["mix_labels"] = mix_labels
+            if self.dataset.rect and isinstance(transform, LetterBox):
+                transform.new_shape = self.dataset.batch_shapes[self.dataset.batch[index]]
+            labels = transform(labels)
+            if "mix_labels" in labels:
+                labels.pop("mix_labels")
+        return labels

ultralytics/yolo/data/utils.py

@@ -0,0 +1,177 @@
+import contextlib
+import hashlib
+import os
+
+import cv2
+import numpy as np
+from PIL import ExifTags, Image, ImageOps
+
+from ..utils.general import segments2boxes
+
+HELP_URL = "See https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data"
+IMG_FORMATS = "bmp", "dng", "jpeg", "jpg", "mpo", "png", "tif", "tiff", "webp", "pfm"  # include image suffixes
+VID_FORMATS = "asf", "avi", "gif", "m4v", "mkv", "mov", "mp4", "mpeg", "mpg", "ts", "wmv"  # include video suffixes
+BAR_FORMAT = "{l_bar}{bar:10}{r_bar}{bar:-10b}"  # tqdm bar format
+LOCAL_RANK = int(os.getenv("LOCAL_RANK", -1))  # https://pytorch.org/docs/stable/elastic/run.html
+RANK = int(os.getenv('RANK', -1))
+PIN_MEMORY = str(os.getenv("PIN_MEMORY", True)).lower() == "true"  # global pin_memory for dataloaders
+IMAGENET_MEAN = 0.485, 0.456, 0.406  # RGB mean
+IMAGENET_STD = 0.229, 0.224, 0.225  # RGB standard deviation
+
+# Get orientation exif tag
+for orientation in ExifTags.TAGS.keys():
+    if ExifTags.TAGS[orientation] == "Orientation":
+        break
+
+
+def img2label_paths(img_paths):
+    # Define label paths as a function of image paths
+    sa, sb = f"{os.sep}images{os.sep}", f"{os.sep}labels{os.sep}"  # /images/, /labels/ substrings
+    return [sb.join(x.rsplit(sa, 1)).rsplit(".", 1)[0] + ".txt" for x in img_paths]
+
+
+def get_hash(paths):
+    # Returns a single hash value of a list of paths (files or dirs)
+    size = sum(os.path.getsize(p) for p in paths if os.path.exists(p))  # sizes
+    h = hashlib.md5(str(size).encode())  # hash sizes
+    h.update("".join(paths).encode())  # hash paths
+    return h.hexdigest()  # return hash
+
+
+def exif_size(img):
+    # Returns exif-corrected PIL size
+    s = img.size  # (width, height)
+    with contextlib.suppress(Exception):
+        rotation = dict(img._getexif().items())[orientation]
+        if rotation in [6, 8]:  # rotation 270 or 90
+            s = (s[1], s[0])
+    return s
+
+
+def verify_image_label(args):
+    # Verify one image-label pair
+    im_file, lb_file, prefix, keypoint = args
+    nm, nf, ne, nc, msg, segments, keypoints = 0, 0, 0, 0, "", None, None  # number (missing, found, empty, corrupt), message, segments, keypoints
+    try:
+        # verify images
+        im = Image.open(im_file)
+        im.verify()  # PIL verify
+        shape = exif_size(im)  # image size
+        shape = (shape[1], shape[0])  # hw
+        assert (shape[0] > 9) & (shape[1] > 9), f"image size {shape} <10 pixels"
+        assert im.format.lower() in IMG_FORMATS, f"invalid image format {im.format}"
+        if im.format.lower() in ("jpg", "jpeg"):
+            with open(im_file, "rb") as f:
+                f.seek(-2, 2)
+                if f.read() != b"\xff\xd9":  # corrupt JPEG
+                    ImageOps.exif_transpose(Image.open(im_file)).save(im_file, "JPEG", subsampling=0, quality=100)
+                    msg = f"{prefix}WARNING ⚠️ {im_file}: corrupt JPEG restored and saved"
+
+        # verify labels
+        if os.path.isfile(lb_file):
+            nf = 1  # label found
+            with open(lb_file) as f:
+                lb = [x.split() for x in f.read().strip().splitlines() if len(x)]
+                if any(len(x) > 6 for x in lb) and (not keypoint):  # is segment
+                    classes = np.array([x[0] for x in lb], dtype=np.float32)
+                    segments = [np.array(x[1:], dtype=np.float32).reshape(-1, 2) for x in lb]  # (cls, xy1...)
+                    lb = np.concatenate((classes.reshape(-1, 1), segments2boxes(segments)), 1)  # (cls, xywh)
+                lb = np.array(lb, dtype=np.float32)
+            nl = len(lb)
+            if nl:
+                if keypoint:
+                    assert lb.shape[1] == 56, "labels require 56 columns each"
+                    assert (lb[:, 5::3] <= 1).all(), "non-normalized or out of bounds coordinate labels"
+                    assert (lb[:, 6::3] <= 1).all(), "non-normalized or out of bounds coordinate labels"
+                    kpts = np.zeros((lb.shape[0], 39))
+                    for i in range(len(lb)):
+                        kpt = np.delete(lb[i, 5:], np.arange(2, lb.shape[1] - 5,
+                                                             3))  # remove the occlusion paramater from the GT
+                        kpts[i] = np.hstack((lb[i, :5], kpt))
+                    lb = kpts
+                    assert lb.shape[1] == 39, "labels require 39 columns each after removing occlusion paramater"
+                else:
+                    assert lb.shape[1] == 5, f"labels require 5 columns, {lb.shape[1]} columns detected"
+                    assert (lb >= 0).all(), f"negative label values {lb[lb < 0]}"
+                    assert (lb[:, 1:] <=
+                            1).all(), f"non-normalized or out of bounds coordinates {lb[:, 1:][lb[:, 1:] > 1]}"
+                _, i = np.unique(lb, axis=0, return_index=True)
+                if len(i) < nl:  # duplicate row check
+                    lb = lb[i]  # remove duplicates
+                    if segments:
+                        segments = [segments[x] for x in i]
+                    msg = f"{prefix}WARNING ⚠️ {im_file}: {nl - len(i)} duplicate labels removed"
+            else:
+                ne = 1  # label empty
+                lb = np.zeros((0, 39), dtype=np.float32) if keypoint else np.zeros((0, 5), dtype=np.float32)
+        else:
+            nm = 1  # label missing
+            lb = np.zeros((0, 39), dtype=np.float32) if keypoint else np.zeros((0, 5), dtype=np.float32)
+        if keypoint:
+            keypoints = lb[:, 5:].reshape(-1, 17, 2)
+        lb = lb[:, :5]
+        return im_file, lb, shape, segments, keypoints, nm, nf, ne, nc, msg
+    except Exception as e:
+        nc = 1
+        msg = f"{prefix}WARNING ⚠️ {im_file}: ignoring corrupt image/label: {e}"
+        return [None, None, None, None, None, nm, nf, ne, nc, msg]
+
+
+def polygon2mask(img_size, polygons, color=1, downsample_ratio=1):
+    """
+    Args:
+        img_size (tuple): The image size.
+        polygons (np.ndarray): [N, M], N is the number of polygons,
+            M is the number of points(Be divided by 2).
+    """
+    mask = np.zeros(img_size, dtype=np.uint8)
+    polygons = np.asarray(polygons)
+    polygons = polygons.astype(np.int32)
+    shape = polygons.shape
+    polygons = polygons.reshape(shape[0], -1, 2)
+    cv2.fillPoly(mask, polygons, color=color)
+    nh, nw = (img_size[0] // downsample_ratio, img_size[1] // downsample_ratio)
+    # NOTE: fillPoly firstly then resize is trying the keep the same way
+    # of loss calculation when mask-ratio=1.
+    mask = cv2.resize(mask, (nw, nh))
+    return mask
+
+
+def polygons2masks(img_size, polygons, color, downsample_ratio=1):
+    """
+    Args:
+        img_size (tuple): The image size.
+        polygons (list[np.ndarray]): each polygon is [N, M],
+            N is the number of polygons,
+            M is the number of points(Be divided by 2).
+    """
+    masks = []
+    for si in range(len(polygons)):
+        mask = polygon2mask(img_size, [polygons[si].reshape(-1)], color, downsample_ratio)
+        masks.append(mask)
+    return np.array(masks)
+
+
+def polygons2masks_overlap(img_size, segments, downsample_ratio=1):
+    """Return a (640, 640) overlap mask."""
+    masks = np.zeros((img_size[0] // downsample_ratio, img_size[1] // downsample_ratio),
+                     dtype=np.int32 if len(segments) > 255 else np.uint8)
+    areas = []
+    ms = []
+    for si in range(len(segments)):
+        mask = polygon2mask(
+            img_size,
+            [segments[si].reshape(-1)],
+            downsample_ratio=downsample_ratio,
+            color=1,
+        )
+        ms.append(mask)
+        areas.append(mask.sum())
+    areas = np.asarray(areas)
+    index = np.argsort(-areas)
+    ms = np.array(ms)[index]
+    for i in range(len(segments)):
+        mask = ms[i] * (i + 1)
+        masks = masks + mask
+        masks = np.clip(masks, a_min=0, a_max=i + 1)
+    return masks, index