sharpiless / yolov5-deepsort-inference Goto Github PK

class Detector(baseDet):

    def __init__(self):
        super(Detector, self).__init__()
        self.init_model()
        self.build_config()

    def init_model(self):

        self.weights = 'weights/yolov5m.pt'
        self.device = '0' if torch.cuda.is_available() else 'cpu'
        self.device = select_device(self.device)
        model = attempt_load(self.weights, map_location=self.device)
        model.to(self.device).eval()
        model.half()
        # torch.save(model, 'test.pt')
        self.m = model
        self.names = model.module.names if hasattr(
            model, 'module') else model.names

    def preprocess(self, img):

        img0 = img.copy()
        img = letterbox(img, new_shape=self.img_size)[0]
        img = img[:, :, ::-1].transpose(2, 0, 1)
        img = np.ascontiguousarray(img)
        img = torch.from_numpy(img).to(self.device)
        img = img.half()  # 半精度
        img /= 255.0  # 图像归一化
        if img.ndimension() == 3:
            img = img.unsqueeze(0)

        return img0, img

    def detect(self, im):

        im0, img = self.preprocess(im)

        pred = self.m(img, augment=False)[0]
        pred = pred.float()
        pred = non_max_suppression(pred, self.threshold, 0.4)

        pred_boxes = []
        for det in pred:

            if det is not None and len(det):
                det[:, :4] = scale_coords(
                    img.shape[2:], det[:, :4], im0.shape).round()

                for *x, conf, cls_id in det:
                    lbl = self.names[int(cls_id)]
                    if not lbl in ['person', 'car', 'truck']:
                        continue
                    x1, y1 = int(x[0]), int(x[1])
                    x2, y2 = int(x[2]), int(x[3])
                    pred_boxes.append(
                        (x1, y1, x2, y2, lbl, conf))

        return im, pred_boxes

调用 self.detect 方法返回图像和预测结果

DeepSort追踪器：

deepsort = DeepSort(cfg.DEEPSORT.REID_CKPT,
                    max_dist=cfg.DEEPSORT.MAX_DIST, min_confidence=cfg.DEEPSORT.MIN_CONFIDENCE,
                    nms_max_overlap=cfg.DEEPSORT.NMS_MAX_OVERLAP, max_iou_distance=cfg.DEEPSORT.MAX_IOU_DISTANCE,
                    max_age=cfg.DEEPSORT.MAX_AGE, n_init=cfg.DEEPSORT.N_INIT, nn_budget=cfg.DEEPSORT.NN_BUDGET,
                    use_cuda=True)

调用 self.update 方法更新追踪结果

运行demo：

python demo.py

训练自己的模型：

参考我的另一篇博客：

【小白CV】手把手教你用YOLOv5训练自己的数据集（从Windows环境配置到模型部署）

训练好后放到 weights 文件夹下

调用接口：

创建检测器：

from AIDetector_pytorch import Detector

det = Detector()

调用检测接口：

func_status = {}
func_status['headpose'] = None

result = det.feedCap(im, func_status)

其中 im 为 BGR 图像

返回的 result 是字典，result['frame'] 返回可视化后的图像

关注我的公众号：

感兴趣的同学关注我的公众号——可达鸭的深度学习教程：

联系作者：

B站：https://space.bilibili.com/470550823

CSDN：https://blog.csdn.net/weixin_44936889

AI Studio：https://aistudio.baidu.com/aistudio/personalcenter/thirdview/67156

Github：https://github.com/Sharpiless

遵循 GNU General Public License v3.0 协议，标明目标检测部分来源：https://github.com/ultralytics/yolov5/

yolov5-deepsort-inference's People

Contributors

Stargazers

Watchers

Forkers

lllovage 591317622a porterpan dy033 angelsu21 hererookie tiqq111 ablek jeremy8080 xialibing g921002 asi-sx yang-xueyi bingweiyang harold-lkk 1085437068 sxjpage wxy1234567 abcd19886 liuxiaoxiao666 microengine yangyin2016 wulongyuan ivan-vv zhangmingjiang1114 onlyjeep zhaoxiaolong2020 ysour freewind2016 smallflyfly xiaojunzaizai kevinxu816 yangliang3 todeem hsueeh wangyunfan vincent-winner tonight123 rockwong66 quliulangle ynt233 feibazhf w642833823 jiehwa xiangyong lichongbing chunde quanh1990 stanwhy xxyqsy zhengjieyuan yfq512 027kaka collinluo euui joez1015 justinlxf huangxianmin888 ishida83 andytyc liupig casually blue-cosmos 782309745 jojinshallar caojunhai ruslanyy aifact pikpiksec pjdu wangwei7911 hnghng matols kato1013 wujimin515600 wanghuisenior ranchogoose xinhong882 mmdwjsj zjj-2015 xwlchengxu dddd-shwy fengshaowei0219 jiancheng-ai lihuaqiang0101 feng62 huigs jackerho tingwenypx dramah aexleader liuhk123 freet15 zhangtianqing hizengbiao beimu6666 knightofdawn zhiweidaxia lzdimu liulongge

yolov5-deepsort-inference's Issues

只有cpu的话是不是跑不了demo

运行报错 RuntimeError: "slow_conv2d_cpu" not implemented for 'Half'
查了下是Half这个变量没法在cpu上跑，所以想问下win10+cpu是不是无法泡桐该项目

Traceback (most recent call last):
File "E:/TuXiangChuLi/week14/Yolov5-deepsort-inference/demo.py", line 1, in
from AIDetector_pytorch import Detector
File "E:\TuXiangChuLi\week14\Yolov5-deepsort-inference\AIDetector_pytorch.py", line 6, in
from utils.BaseDetector import baseDet
File "E:\TuXiangChuLi\week14\Yolov5-deepsort-inference\utils\BaseDetector.py", line 1, in
from tracker import update_tracker
File "E:\TuXiangChuLi\week14\Yolov5-deepsort-inference\tracker.py", line 8, in
cfg.merge_from_file("E:/TuXiangChuLi/week14/Yolov5-deepsort-inference/deep_sort/configs/deep_sort.yaml")
File "E:\TuXiangChuLi\week14\Yolov5-deepsort-inference\deep_sort\utils\parser.py", line 23, in merge_from_file
self.update(yaml.load(fo.read()))
TypeError: load() missing 1 required positional argument: 'Loader'

Process finished with exit code 1

feature request : adding tracking tail

作者您知道怎么加上那个跟踪轨迹的"尾巴"吗?(不知道您理解我说的这个尾巴吗)后面会加上吗,谢谢!

请问哪里有呢c++版本的deepsort？求助

how to train the deep sort model?

hello! thanks for your open source.
but i want to know that how to train the deep sort model which can used in your repo?
Could you give a instruction to us?
THX

出现too many indices for tensor of dimension 1，请问这个问题该怎么解决呢

关与Box格式的问题

tracker.py中和deep_sort.py中对bbox的xyxy和xywh定义不一

能否用于自定义的数据集

我已经用自定义的数据集训练了yolov5的模型，请问用deepsort还要再训练相应的模型权重么？训练的数据集格式是什么？

一个小问题

ERROR: Could not open requirements file: [Errno 2] No such file or directory: 'requirements.txt'
文件里没有这个 requirements.txt 啊，我就装不上了

只能使用yolov5m吗

demo中是不是只有识别没有跟踪？

追踪部分太耗cpu

hello，追踪这部分哪里可以修改的吗，cpu顶不住，占用太多

error

RuntimeError: "unfolded2d_copy" not implemented for 'Half'

请问deep sort的模型是在车辆和行人的数据集上进行了重新训练吗？

当视频中没有人或者车可以检测时就报错

RuntimeError: "upsample_nearest2d_channels_last" not implemented for 'Half'

我上網查了，好像跟torch版本有關，但現在只能裝2.x版而已，請問一下該如何解決?
Traceback (most recent call last):
File "c:\yolov5_deepsort_infer\Yolov5-deepsort-inference\demo.py", line 58, in
main()
File "c:\yolov5_deepsort_infer\Yolov5-deepsort-inference\demo.py", line 32, in main
result = det.feedCap(im, func_status)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "c:\yolov5_deepsort_infer\Yolov5-deepsort-inference\utils\BaseDetector.py", line 35, in feedCap
im, faces, face_bboxes = update_tracker(self, im)
^^^^^^^^^^^^^^^^^^^^^^^^
File "c:\yolov5_deepsort_infer\Yolov5-deepsort-inference\tracker.py", line 42, in update_tracker
_, bboxes = target_detector.detect(image)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "c:\yolov5_deepsort_infer\Yolov5-deepsort-inference\AIDetector_pytorch.py", line 47, in detect
pred = self.m(img, augment=False)[0]
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\yolov5_deepsort_infer\yolov5_infer\Lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\yolov5_deepsort_infer\yolov5_infer\Lib\site-packages\torch\nn\modules\module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "c:\yolov5_deepsort_infer\Yolov5-deepsort-inference\models\yolo.py", line 112, in forward
return self.forward_once(x, profile) # single-scale inference, train
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "c:\yolov5_deepsort_infer\Yolov5-deepsort-inference\models\yolo.py", line 128, in forward_once
x = m(x) # run
^^^^
File "C:\yolov5_deepsort_infer\yolov5_infer\Lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\yolov5_deepsort_infer\yolov5_infer\Lib\site-packages\torch\nn\modules\module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\yolov5_deepsort_infer\yolov5_infer\Lib\site-packages\torch\nn\modules\upsampling.py", line 157, in forward
return F.interpolate(input, self.size, self.scale_factor, self.mode, self.align_corners,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\yolov5_deepsort_infer\yolov5_infer\Lib\site-packages\torch\nn\functional.py", line 4001, in interpolate
return torch._C._nn.upsample_nearest2d(input, output_size, scale_factors)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: "upsample_nearest2d_channels_last" not implemented for 'Half'

您解决这个问题了吗？我遇到了相同的问题，如能解答，万分感谢！

同機器使用官方yolov5m檢測時速度與官方給的執行速度也差很多

作者有發現這問題嘛？

謝謝

报错

This application failed to start because no Qt platform plugin could be initialized. Reinstalling the application may fix this problem.

算法评估

如何评估算法性能，是否有相关的代码

sharpiless / yolov5-deepsort-inference Goto Github PK

yolov5-deepsort-inference's Introduction

注意：

本项目使用Yolov5 3.0版本，最新版本5.0请移步：

注：新版本添加了类别显示功能

项目简介：

YOLOv5检测器：