关于推理图片的问题

[Sunshulong]

作者你好，
现在只有rtdetr_paddle里有infer程序，在torch版本里如何进行推断呢？还有paddle和torch训练出来的精度一样吗

[lyuwenyu]

• 推理的话需要自己写前后处理 参考 RTDETR Deployment #95
• 精度一致的 参考pytorch rtdetr_r50 log Pytorch RTDETR #42

[qgq99]

作者你好， 现在只有rtdetr_paddle里有infer程序，在torch版本里如何进行推断呢？还有paddle和torch训练出来的精度一样吗

我参考作者https://github.com/lyuwenyu/RT-DETR/blob/main/rtdetr_pytorch/tools/export_onnx.py 这个文件写了一份torch的推理，可供你参考：

import torch
from torch import nn
from torchvision.transforms import transforms
from PIL import Image, ImageDraw
import sys
sys.path.append("..")
from src.core import YAMLConfig
import argparse
from pathlib import Path
import time


class ImageReader:
    def __init__(self, resize=224, mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]):
        self.transform = transforms.Compose([
            # transforms.Resize((resize, resize)) if isinstance(resize, int) else transforms.Resize(
            #     (resize[0], resize[1])),
            transforms.ToTensor(),
            # transforms.Normalize(mean=mean, std=std),
        ])
        self.resize = resize
        self.pil_img = None   #保存最近一次读取的图片的pil对象

    def __call__(self, image_path, *args, **kwargs):
        """
        读取图片
        """
        self.pil_img = Image.open(image_path).convert('RGB').resize((self.resize, self.resize))
        return self.transform(self.pil_img).unsqueeze(0)




class Model(nn.Module):
    def __init__(self, confg=None, ckpt="") -> None:
        super().__init__()
        self.cfg = YAMLConfig(confg, resume=ckpt)
        if ckpt:
            checkpoint = torch.load(ckpt, map_location='cpu') 
            if 'ema' in checkpoint:
                state = checkpoint['ema']['module']
            else:
                state = checkpoint['model']
        else:
            raise AttributeError('only support resume to load model.state_dict by now.')

        # NOTE load train mode state -> convert to deploy mode
        self.cfg.model.load_state_dict(state)

        self.model = self.cfg.model.deploy()
        self.postprocessor = self.cfg.postprocessor.deploy()
        # print(self.postprocessor.deploy_mode)
        
    def forward(self, images, orig_target_sizes):
        outputs = self.model(images)
        return self.postprocessor(outputs, orig_target_sizes)



def get_argparser():
    parser = argparse.ArgumentParser()
    parser.add_argument("--config", default="rt-DETR/rtdetr_pytorch/configs/rtdetr/rtdetr_r50vd_6x_coco.yml", help="配置文件路径")
    parser.add_argument("--ckpt", default="rt-DETR/rtdetr_pytorch/weights/rtdetr_r50vd_6x_coco_from_paddle.pth", help="权重文件路径")
    parser.add_argument("--image", default="rt-DETR/rtdetr_pytorch/images/keji.jpg", help="待推理图片路径")
    parser.add_argument("--output_dir", default="rt-DETR/rtdetr_pytorch/images/ouput", help="输出文件保存路径")
    parser.add_argument("--device", default="cpu")

    return parser


def main(args):
    img_path = Path(args.image)
    device = torch.device(args.device)
    reader = ImageReader(resize=640)
    model = Model(confg=args.config, ckpt=args.ckpt)
    model.to(device=device)

    img =reader(img_path).to(device)
    size = torch.tensor([[img.shape[2], img.shape[3]]]).to(device)
    start = time.time()
    output = model(img, size)
    print(f"推理耗时：{time.time() - start:.4f}s")
    labels, boxes, scores = output
    im = reader.pil_img
    draw = ImageDraw.Draw(im)
    thrh = 0.6

    for i in range(img.shape[0]):

        scr = scores[i]
        lab = labels[i][scr > thrh]
        box = boxes[i][scr > thrh]


        for b in box:
            draw.rectangle(list(b), outline='red', )
            draw.text((b[0], b[1]), text=str(lab[i]), fill='blue', )

    save_path = Path(args.output_dir) / img_path.name
    im.save(save_path)
    print(f"检测结果已保存至:{save_path}")

if __name__ == "__main__":
    main(get_argparser().parse_args())
    

[leejq666]

作者你好， 现在只有rtdetr_paddle里有infer程序，在torch版本里如何进行推断呢？还有paddle和torch训练出来的精度一样吗

我参考作者https://github.com/lyuwenyu/RT-DETR/blob/main/rtdetr_pytorch/tools/export_onnx.py 这个文件写了一份torch的推理，可供你参考：

import torch
from torch import nn
from torchvision.transforms import transforms
from PIL import Image, ImageDraw
import sys
sys.path.append("..")
from src.core import YAMLConfig
import argparse
from pathlib import Path
import time


class ImageReader:
    def __init__(self, resize=224, mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]):
        self.transform = transforms.Compose([
            # transforms.Resize((resize, resize)) if isinstance(resize, int) else transforms.Resize(
            #     (resize[0], resize[1])),
            transforms.ToTensor(),
            # transforms.Normalize(mean=mean, std=std),
        ])
        self.resize = resize
        self.pil_img = None   #保存最近一次读取的图片的pil对象

    def __call__(self, image_path, *args, **kwargs):
        """
        读取图片
        """
        self.pil_img = Image.open(image_path).convert('RGB').resize((self.resize, self.resize))
        return self.transform(self.pil_img).unsqueeze(0)




class Model(nn.Module):
    def __init__(self, confg=None, ckpt="") -> None:
        super().__init__()
        self.cfg = YAMLConfig(confg, resume=ckpt)
        if ckpt:
            checkpoint = torch.load(ckpt, map_location='cpu') 
            if 'ema' in checkpoint:
                state = checkpoint['ema']['module']
            else:
                state = checkpoint['model']
        else:
            raise AttributeError('only support resume to load model.state_dict by now.')

        # NOTE load train mode state -> convert to deploy mode
        self.cfg.model.load_state_dict(state)

        self.model = self.cfg.model.deploy()
        self.postprocessor = self.cfg.postprocessor.deploy()
        # print(self.postprocessor.deploy_mode)
        
    def forward(self, images, orig_target_sizes):
        outputs = self.model(images)
        return self.postprocessor(outputs, orig_target_sizes)



def get_argparser():
    parser = argparse.ArgumentParser()
    parser.add_argument("--config", default="rt-DETR/rtdetr_pytorch/configs/rtdetr/rtdetr_r50vd_6x_coco.yml", help="配置文件路径")
    parser.add_argument("--ckpt", default="rt-DETR/rtdetr_pytorch/weights/rtdetr_r50vd_6x_coco_from_paddle.pth", help="权重文件路径")
    parser.add_argument("--image", default="rt-DETR/rtdetr_pytorch/images/keji.jpg", help="待推理图片路径")
    parser.add_argument("--output_dir", default="rt-DETR/rtdetr_pytorch/images/ouput", help="输出文件保存路径")
    parser.add_argument("--device", default="cpu")

    return parser


def main(args):
    img_path = Path(args.image)
    device = torch.device(args.device)
    reader = ImageReader(resize=640)
    model = Model(confg=args.config, ckpt=args.ckpt)
    model.to(device=device)

    img =reader(img_path).to(device)
    size = torch.tensor([[img.shape[2], img.shape[3]]]).to(device)
    start = time.time()
    output = model(img, size)
    print(f"推理耗时：{time.time() - start:.4f}s")
    labels, boxes, scores = output
    im = reader.pil_img
    draw = ImageDraw.Draw(im)
    thrh = 0.6

    for i in range(img.shape[0]):

        scr = scores[i]
        lab = labels[i][scr > thrh]
        box = boxes[i][scr > thrh]


        for b in box:
            draw.rectangle(list(b), outline='red', )
            draw.text((b[0], b[1]), text=str(lab[i]), fill='blue', )

    save_path = Path(args.output_dir) / img_path.name
    im.save(save_path)
    print(f"检测结果已保存至:{save_path}")

if __name__ == "__main__":
    main(get_argparser().parse_args())
    

您好！非常感谢您的分享，对我的帮助很大！：）我这里可视化结果方框上面的不是类别和对应的概率，而是device.对于这样的情况您有相应的解决方案吗？非常期待您的回复！

[zhang123-lf]

作者你好，现在只有rtdetr_paddle里有infer程序，在torch版本里如何进行推断呢？还有paddle和torch训练出来的精度一样吗？

我参考作者https://github.com/lyuwenyu/RT-DETR/blob/main/rtdetr_pytorch/tools/export_onnx.py这个文件写了一份torch的推理，供你参考：

import torch
from torch import nn
from torchvision.transforms import transforms
from PIL import Image, ImageDraw
import sys
sys.path.append("..")
from src.core import YAMLConfig
import argparse
from pathlib import Path
import time

class ImageReader:
def init(self, resize=224, mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]):
self.transform = transforms.Compose([
# transforms.Resize((resize, resize)) if isinstance(resize, int) else transforms.Resize(
# (resize[0], resize[1])),
transforms.ToTensor(),
# transforms.Normalize(mean=mean, std=std),
])
self.resize = resize
self.pil_img = None #保存最近一次读取的图片的pil对象

def __call__(self, image_path, *args, **kwargs):
    """
    读取图片
    """
    self.pil_img = Image.open(image_path).convert('RGB').resize((self.resize, self.resize))
    return self.transform(self.pil_img).unsqueeze(0)

class Model(nn.Module):
def init(self, confg=None, ckpt="") -> None:
super().init()
self.cfg = YAMLConfig(confg, resume=ckpt)
if ckpt:
checkpoint = torch.load(ckpt, map_location='cpu')
if 'ema' in checkpoint:
state = checkpoint['ema']['module']
else:
state = checkpoint['model']
else:
raise AttributeError('only support resume to load model.state_dict by now.')

    # NOTE load train mode state -> convert to deploy mode
    self.cfg.model.load_state_dict(state)

    self.model = self.cfg.model.deploy()
    self.postprocessor = self.cfg.postprocessor.deploy()
    # print(self.postprocessor.deploy_mode)
    
def forward(self, images, orig_target_sizes):
    outputs = self.model(images)
    return self.postprocessor(outputs, orig_target_sizes)

def get_argparser():
parser = argparse.ArgumentParser()
parser.add_argument("--config", default="rt-DETR/rtdetr_pytorch/configs/rtdetr/rtdetr_r50vd_6x_coco.yml", help="配置文件路径")
parser.add_argument("--ckpt", default="rt-DETR/rtdetr_pytorch/weights/rtdetr_r50vd_6x_coco_from_paddle.pth", help="权重文件路径")
parser.add_argument("--image", default="rt-DETR/rtdetr_pytorch/images/keji.jpg", help="待推理图片路径")
parser.add_argument("--output_dir", default="rt-DETR/rtdetr_pytorch/images/ouput", help="输出文件保存路径")
parser.add_argument("--device", default="cpu")

return parser

def main(args):
img_path = Path(args.image)
device = torch.device(args.device)
reader = ImageReader(resize=640)
model = Model(confg=args.config, ckpt=args.ckpt)
model.to(device=device)

img =reader(img_path).to(device)
size = torch.tensor([[img.shape[2], img.shape[3]]]).to(device)
start = time.time()
output = model(img, size)
print(f"推理耗时：{time.time() - start:.4f}s")
labels, boxes, scores = output
im = reader.pil_img
draw = ImageDraw.Draw(im)
thrh = 0.6

for i in range(img.shape[0]):

    scr = scores[i]
    lab = labels[i][scr > thrh]
    box = boxes[i][scr > thrh]


    for b in box:
        draw.rectangle(list(b), outline='red', )
        draw.text((b[0], b[1]), text=str(lab[i]), fill='blue', )

save_path = Path(args.output_dir) / img_path.name
im.save(save_path)
print(f"检测结果已保存至:{save_path}")

if name == "main":
main(get_argparser().parse_args())

非常感谢你的分享，但是我在GPU上推理一张图片比CPU上的耗时长很多，请问你的推时间是多少