YoloV4 train

classes_path = 'train_parameters/classes.txt'
anchors_path = 'train_parameters/yolo_anchors.txt'
train_anno_path = 'train_parameters/train_anno.txt'
Cuda = True
Use_Data_Loader = True
normalize = False
input_shape = (608,608)
mosaic = False
Cosine_lr = True
smoooth_label = 0
lr = 5e-4
Batch_size = 12
Init_Epoch = 0
Freeze_Epoch = 1000 #总共epoch次数

if __name__ == '__main__':
    #                         model parameters\config
    class_names = get_classes(classes_path)
    anchors = get_anchors(anchors_path)
    num_classes = len(class_names)
    model = YoloBody(len(anchors[0]), num_classes)
    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
    net = model.train()
    if Cuda:
        net = net.cuda()
    yolo_losses = []
    for i in range(3):
        yolo_losses.append(YOLOLoss(np.reshape(anchors, [-1, 2]), num_classes, \
                                    (input_shape[1], input_shape[0]), smoooth_label, Cuda, normalize))

    with open(train_anno_path) as f:
        lines = f.readlines()

    val_split = 0.1
    np.random.seed(10101)
    np.random.shuffle(lines)
    np.random.seed(None)
    num_val = int(len(lines) * val_split * 0.4 )
    num_train = int((len(lines) - num_val) * 0.4 )

    # if True:

    optimizer = optim.Adam(net.parameters(),lr)

    if Cosine_lr:
        lr_scheduler = optim.lr_scheduler.CosineAnnealingLR(optimizer, T_max=5, eta_min=1e-5)
    else:
        lr_scheduler = optim.lr_scheduler.StepLR(optimizer,step_size=1,gamma=1)

    train_dataset = YoloDataset(lines[:num_train], (input_shape[0], input_shape[1]), mosaic=mosaic, is_train=True)
    val_dataset = YoloDataset(lines[num_train:num_train+num_val], (input_shape[0], input_shape[1]), mosaic=False, is_train=False)
    gen = DataLoader(train_dataset, shuffle=True, batch_size=Batch_size, num_workers=0, pin_memory=True,
                            drop_last=True, collate_fn=yolo_dataset_collate)
    gen_val = DataLoader(val_dataset, shuffle=True, batch_size=Batch_size, num_workers=0,pin_memory=True,
                            drop_last=True, collate_fn=yolo_dataset_collate)

    epoch_size = max(1, num_train//Batch_size) # epoch
    epoch_size_val = num_val//Batch_size

    for param in model.backbone.parameters():
        param.requires_grad = False

    for epoch in range(Init_Epoch,Freeze_Epoch):
        fit_one_epoch(net,yolo_losses,epoch,epoch_size,epoch_size_val,gen,gen_val,Freeze_Epoch,Cuda)
        lr_scheduler.step()

from random import shuffle
import numpy as np
import torch
import torch.nn as nn
import math
import torch.nn.functional as F
from PIL import Image
from torch.autograd import Variable
from torch.utils.data import DataLoader
from torch.utils.data.dataset import Dataset
from utils.utils import bbox_iou, merge_bboxes
from matplotlib.colors import rgb_to_hsv, hsv_to_rgb
from nets.yolo_training import Generator
import cv2
import copy

class YoloDataset(Dataset):
    def __init__(self, train_lines, image_size, mosaic=True, is_train=True):
        super(YoloDataset, self).__init__()

        self.train_lines = train_lines
        self.train_batches = len(train_lines)
        self.image_size = image_size
        self.mosaic = mosaic
        self.flag = True
        self.is_train = is_train

    def __len__(self):
        return self.train_batches

    def rand(self, a=0, b=1):
        return np.random.rand() * (b - a) + a

    def get_random_data(self, annotation_line, input_shape, jitter=.3, hue=.1, sat=1.5, val=1.5, random=True):
        """实时数据增强的随机预处理"""
        if '.jpg' in annotation_line:
            img_path = annotation_line.split(".jpg",1)[0]+'.jpg'
            line = copy.deepcopy(annotation_line.split(".jpg", 1)[1].split())
        else:
            img_path = annotation_line.split(".png", 1)[0] + '.png'
            line = copy.deepcopy(annotation_line.split(".png", 1)[1].split())
        line.insert(0,img_path)
        image = Image.open(line[0])
        iw, ih = image.size
        h, w = input_shape
        box = np.array([np.array(list(map(int, box.split(',')))) for box in line[1:]])

        if not random:
            scale = min(w/iw, h/ih)
            nw = int(iw*scale)
            nh = int(ih*scale)
            dx = (w-nw)//2
            dy = (h-nh)//2

            image = image.resize((nw,nh), Image.BICUBIC)
            new_image = Image.new('RGB', (w,h), (128,128,128))
            new_image.paste(image, (dx, dy))
            image_data = np.array(new_image, np.float32)

            # 调整目标框坐标
            box_data = np.zeros((len(box), 5))
            if len(box) > 0:
                np.random.shuffle(box)
                box[:, [0, 2]] = box[:, [0, 2]] * nw / iw + dx
                box[:, [1, 3]] = box[:, [1, 3]] * nh / ih + dy
                box[:, 0:2][box[:, 0:2] < 0] = 0
                box[:, 2][box[:, 2] > w] = w
                box[:, 3][box[:, 3] > h] = h
                box_w = box[:, 2] - box[:, 0]
                box_h = box[:, 3] - box[:, 1]
                box = box[np.logical_and(box_w > 1, box_h > 1)]  # 保留有效框
                box_data = np.zeros((len(box), 5))
                box_data[:len(box)] = box

            return image_data, box_data

        # 调整图片大小
        new_ar = w / h * self.rand(1 - jitter, 1 + jitter) / self.rand(1 - jitter, 1 + jitter) #
        scale = self.rand(.25, 2) # 0.25-2.0 缩放倍数范围
        if new_ar < 1:
            nh = int(scale * h)
            nw = int(nh * new_ar)
        else:
            nw = int(scale * w)
            nh = int(nw / new_ar)
        image = image.resize((nw, nh), Image.BICUBIC)

        # 放置图片
        dx = int(self.rand(0, w - nw))
        dy = int(self.rand(0, h - nh))
        new_image = Image.new('RGB', (w, h),
                              (np.random.randint(0, 255), np.random.randint(0, 255), np.random.randint(0, 255)))
        new_image.paste(image, (dx, dy))
        image = new_image

        # 是否翻转图片
        flip = self.rand() < .5
        if flip:
            image = image.transpose(Image.FLIP_LEFT_RIGHT)

        # 色域变换
        hue = self.rand(-hue, hue)
        sat = self.rand(1, sat) if self.rand() < .5 else 1 / self.rand(1, sat)
        val = self.rand(1, val) if self.rand() < .5 else 1 / self.rand(1, val)
        x = cv2.cvtColor(np.array(image,np.float32)/255, cv2.COLOR_RGB2HSV)
        x[..., 0] += hue*360
        x[..., 0][x[..., 0]>1] -= 1
        x[..., 0][x[..., 0]<0] += 1
        x[..., 1] *= sat
        x[..., 2] *= val
        x[x[:,:, 0]>360, 0] = 360
        x[:, :, 1:][x[:, :, 1:]>1] = 1
        x[x<0] = 0
        image_data = cv2.cvtColor(x, cv2.COLOR_HSV2RGB)*255

        # 调整目标框坐标
        box_data = np.zeros((len(box), 5))
        if len(box) > 0:
            np.random.shuffle(box)
            box[:, [0, 2]] = box[:, [0, 2]] * nw / iw + dx
            box[:, [1, 3]] = box[:, [1, 3]] * nh / ih + dy
            if flip:
                box[:, [0, 2]] = w - box[:, [2, 0]]
            box[:, 0:2][box[:, 0:2] < 0] = 0
            box[:, 2][box[:, 2] > w] = w
            box[:, 3][box[:, 3] > h] = h
            box_w = box[:, 2] - box[:, 0]
            box_h = box[:, 3] - box[:, 1]
            box = box[np.logical_and(box_w > 1, box_h > 1)]  # 保留有效框
            box_data = np.zeros((len(box), 5))
            box_data[:len(box)] = box

        return image_data, box_data

    def get_random_data_with_Mosaic(self, annotation_line, input_shape, hue=.1, sat=1.5, val=1.5):
        h, w = input_shape
        min_offset_x = 0.3
        min_offset_y = 0.3
        scale_low = 1 - min(min_offset_x, min_offset_y)
        scale_high = scale_low + 0.2

        image_datas = []
        box_datas = []
        index = 0

        place_x = [0, 0, int(w * min_offset_x), int(w * min_offset_x)]
        place_y = [0, int(h * min_offset_y), int(h * min_offset_y), 0]
        for line in annotation_line:
            # 每一行进行分割
            line_content = line.split()
            # 打开图片
            image = Image.open(line_content[0])
            image = image.convert("RGB")
            # 图片的大小
            iw, ih = image.size
            # 保存框的位置
            box = np.array([np.array(list(map(int, box.split(',')))) for box in line_content[1:]])

            # 是否翻转图片
            flip = self.rand() < .5
            if flip and len(box) > 0:
                image = image.transpose(Image.FLIP_LEFT_RIGHT)
                box[:, [0, 2]] = iw - box[:, [2, 0]]

            # 对输入进来的图片进行缩放
            new_ar = w / h
            scale = self.rand(scale_low, scale_high)
            if new_ar < 1:
                nh = int(scale * h)
                nw = int(nh * new_ar)
            else:
                nw = int(scale * w)
                nh = int(nw / new_ar)
            image = image.resize((nw, nh), Image.BICUBIC)

            # 进行色域变换
            hue = self.rand(-hue, hue)
            sat = self.rand(1, sat) if self.rand() < .5 else 1 / self.rand(1, sat)
            val = self.rand(1, val) if self.rand() < .5 else 1 / self.rand(1, val)
            x = cv2.cvtColor(np.array(image,np.float32)/255, cv2.COLOR_RGB2HSV)
            x[..., 0] += hue*360
            x[..., 0][x[..., 0]>1] -= 1
            x[..., 0][x[..., 0]<0] += 1
            x[..., 1] *= sat
            x[..., 2] *= val
            x[x[:,:, 0]>360, 0] = 360
            x[:, :, 1:][x[:, :, 1:]>1] = 1
            x[x<0] = 0
            image = cv2.cvtColor(x, cv2.COLOR_HSV2RGB) # numpy array, 0 to 1

            image = Image.fromarray((image * 255).astype(np.uint8))
            # 将图片进行放置，分别对应四张分割图片的位置
            dx = place_x[index]
            dy = place_y[index]
            new_image = Image.new('RGB', (w, h),
                                  (np.random.randint(0, 255), np.random.randint(0, 255), np.random.randint(0, 255)))
            new_image.paste(image, (dx, dy))
            image_data = np.array(new_image)

            index = index + 1
            box_data = []
            # 对box进行重新处理
            if len(box) > 0:
                np.random.shuffle(box)
                box[:, [0, 2]] = box[:, [0, 2]] * nw / iw + dx
                box[:, [1, 3]] = box[:, [1, 3]] * nh / ih + dy
                box[:, 0:2][box[:, 0:2] < 0] = 0
                box[:, 2][box[:, 2] > w] = w
                box[:, 3][box[:, 3] > h] = h
                box_w = box[:, 2] - box[:, 0]
                box_h = box[:, 3] - box[:, 1]
                box = box[np.logical_and(box_w > 1, box_h > 1)]
                box_data = np.zeros((len(box), 5))
                box_data[:len(box)] = box

            image_datas.append(image_data)
            box_datas.append(box_data)

        # 将图片分割，放在一起
        cutx = np.random.randint(int(w * min_offset_x), int(w * (1 - min_offset_x)))
        cuty = np.random.randint(int(h * min_offset_y), int(h * (1 - min_offset_y)))

        new_image = np.zeros([h, w, 3])
        new_image[:cuty, :cutx, :] = image_datas[0][:cuty, :cutx, :]
        new_image[cuty:, :cutx, :] = image_datas[1][cuty:, :cutx, :]
        new_image[cuty:, cutx:, :] = image_datas[2][cuty:, cutx:, :]
        new_image[:cuty, cutx:, :] = image_datas[3][:cuty, cutx:, :]

        # 对框进行进一步的处理
        new_boxes = np.array(merge_bboxes(box_datas, cutx, cuty))

        return new_image, new_boxes

    def __getitem__(self, index):
        lines = self.train_lines
        n = self.train_batches
        index = index % n
        if self.mosaic:
            if self.flag and (index + 4) < n:
                img, y = self.get_random_data_with_Mosaic(lines[index:index + 4], self.image_size[0:2])
            else:
                img, y = self.get_random_data(lines[index], self.image_size[0:2], random=self.is_train)
            self.flag = bool(1-self.flag)
        else:
            img, y = self.get_random_data(lines[index], self.image_size[0:2], random=self.is_train)

        if len(y) != 0:
            # 从坐标转换成0~1的百分比
            boxes = np.array(y[:, :4], dtype=np.float32)
            boxes[:, 0] = boxes[:, 0] / self.image_size[1]
            boxes[:, 1] = boxes[:, 1] / self.image_size[0]
            boxes[:, 2] = boxes[:, 2] / self.image_size[1]
            boxes[:, 3] = boxes[:, 3] / self.image_size[0]

            boxes = np.maximum(np.minimum(boxes, 1), 0)
            boxes[:, 2] = boxes[:, 2] - boxes[:, 0]
            boxes[:, 3] = boxes[:, 3] - boxes[:, 1]

            boxes[:, 0] = boxes[:, 0] + boxes[:, 2] / 2
            boxes[:, 1] = boxes[:, 1] + boxes[:, 3] / 2
            y = np.concatenate([boxes, y[:, -1:]], axis=-1)

        img = np.array(img, dtype=np.float32)

        tmp_inp = np.transpose(img / 255.0, (2, 0, 1))
        tmp_targets = np.array(y, dtype=np.float32)
        return tmp_inp, tmp_targets


# DataLoader中collate_fn使用
def yolo_dataset_collate(batch):
    images = []
    bboxes = []
    for img, box in batch:
        images.append(img)
        bboxes.append(box)
    images = np.array(images)
    return images, bboxes