yehengchen / object-detection-and-tracking Goto Github PK

Why is this?

if len(pts[track.track_id]) != None:
UnboundLocalError: local variable 'track' referenced before assignment

师兄你好，YOLOv3 + SORT - Person Counter这个模型是否使用的gpu？

如图，一小段视频用了699s，使用nVidia-smi查看，貌似没有使用gpu，

如何使用gpu加速？？

请问您使用deep sort的时候自己训练了车辆模型吗？能否分享一下，谢谢。

以及deepsort 能否改成weights檔

Hi, thanks for your great project.
In your project "YOLOv3 + Deep_SORT", I find that your first git demo can track the 'car' and 'person' at same time(the first video here). I really want to know how to do that, and would you mind show me this code?

Hi, Many thanks for great project. But i got into trouble by this problem whenever i run this project about 10 seconds :
. I really want to know how to fix it ! Please help me. :(

Hi,

I ran the code of deep_sort_yolov3 with instruction as python3 main.py -c person -i data/cam1.avi
I met the following error which seems to be caused by the shape mismatch.
Do you have any suggestion how to fix this?
And can you provide a testing video?
Thank you.

InvalidArgumentError (see above for traceback): Input to reshape is a tensor with 312375 values, but the requested shape requires a multiple of 40425 [[Node: Reshape_3 = Reshape[T=DT_FLOAT, Tshape=DT_INT32, _device="/job:localhost/replica:0/task:0/device:CPU:0"](conv2d_59/BiasAdd, Reshape_3/shape)]]

Is there any function to know if a vehicle is static or moving?

I could not find good video for testing.

When running the project yolov3+sort, the video cannot be output, and the runtime results are as follows:
ascend@ubuntu:~/ascend/Demo_Yolov3_Sort/yolov3_sort$ python3 main.py --input input/TownCentreXVID.avi --output output/TownCentreXVID_Result.avi --yolo yolo-obj
/home/ascend/.local/lib/python3.5/site-packages/sklearn/utils/linear_assignment_.py:22: FutureWarning: The linear_assignment_ module is deprecated in 0.21 and will be removed from 0.23. Use scipy.optimize.linear_sum_assignment instead.
FutureWarning)
[INFO] loading YOLO from disk...
[INFO] 7500 total frames in video
[ERROR:0] global /io/opencv/modules/videoio/src/cap.cpp (415) open VIDEOIO(CV_IMAGES): raised OpenCV exception:

OpenCV(4.2.0) /io/opencv/modules/videoio/src/cap_images.cpp:253: error: (-5:Bad argument) CAP_IMAGES: can't find starting number (in the name of file): output/TownCentreXVID_Result.avi in function 'icvExtractPattern'

[INFO] single frame took 0.5864 seconds
[INFO] estimated total time to finish: 4398.2953
/home/ascend/.local/lib/python3.5/site-packages/sklearn/utils/linear_assignment_.py:128: FutureWarning: The linear_assignment function is deprecated in 0.21 and will be removed from 0.23. Use scipy.optimize.linear_sum_assignment instead.
FutureWarning)

please give me some advice. Thanks.

你好 看见您最新b站投稿，对车辆进行了跟踪计数，我比较困惑的是只需要重新训练车辆yolov3.weights吗 然后更改predicted_class != 'car':吗 这个网址 https://github.com/nwojke/cosine_metric_learning还需要训练吗

After testing my model on this code, the bounding boxes for the objects became odd looking. For example, whenever it detects a car, it shows the box looking like this:

Is there any way I can fix the way the boxes are generated, so that it only surrounds the object?

Another thing I would like to ask about:
How do I display a separate counter for object 'car', and another one for object 'tree'?

Hi yehengchen,
Can you share file 'model_data/market1501.pb'?
Best regards,
PeterPham

How could I track cars and pedestrians at the same time? Exactly like what you got in the demo gifs.
Thank you for reply.

您好:

關於deep_sort_yolov3，我使用2080TI 與E3-1225 [email protected]
實際影片都只能跑十幾的FPS速度，想問您跑的效能如何?

而模型是否可以置換yolov3-tiny

謝謝

I did well with yolov3. But I need more fps.

I wanna use it in deep_sort yolov3.

Converting was well-done. But when I implement main.py then error comes out with below statement.

Traceback (most recent call last):
File "main.py", line 247, in
main(YOLO())
File "/home/user/object/OneStage/yolo/deep_sort_yolov3/yolo.py", line 54, in init
self.boxes, self.scores, self.classes = self.generate()
File "/home/user/object/OneStage/yolo/deep_sort_yolov3/yolo.py", line 94, in generate
score_threshold=self.score, iou_threshold=self.iou)
File "/home/user/object/OneStage/yolo/deep_sort_yolov3/yolo3/model.py", line 169, in yolo_eval
_boxes, _box_scores = yolo_boxes_and_scores(yolo_outputs[l],
IndexError: list index out of range

Could you help me? plz

请问画轨迹线的demo_path,py这个文件在哪里啊？

已经按照以下步骤执行：
Download the code to your computer.

Download [yolov3.weights] and place it in yolov3_sort/yolo-obj/

[yolov3_sort/main.py] Change the Path to labelsPath / weightsPath / configPath. - 更换main.py中的路径

Run the yolov3 counter:

$ python3 main.py --input input/test.mp4 --output output/test.avi --yolo yolo-obj
最终output文件夹里图片均未出框也未计数
我的path:
labelsPath = os.path.sep.join([args["yolo"],"yolo3_object.names"])
weightsPath = os.path.sep.join([args["yolo"],"yolov3.weights"])
configPath = os.path.sep.join([args["yolo"],"yolov3_1.cfg"])
是路径的问题吗？感谢！好人一生平安！

我已经在win10/7上完美的实现效果。但迷惑不知道如何保存检测的视频？求解！！

I also tried other methods. I guess it would be difficult to get it working outside China. Anyone has google drive, mega links of the same? the file name is yolo-spp.h5 and baidu exctraction code is t13k

我是小白刚入门，最近想做个行人检测的项目，如果要用yolo实现您这个项目这样的行人检测，具体步骤是怎样的呢？是修改配置让模型只检测行人就可以了？模型需要修改吗？需要用到迁移学习之类的吗

Traceback (most recent call last):
File "main.py", line 172, in
main(YOLO())
File "/home/xbot/Documents/Object-Detection-and-Tracking-master/OneStage/yolo/deep_sort_yolov3/yolo.py", line 51, in init
self.boxes, self.scores, self.classes = self.generate()
File "/home/xbot/Documents/Object-Detection-and-Tracking-master/OneStage/yolo/deep_sort_yolov3/yolo.py", line 89, in generate
boxes, scores, classes = yolo_eval(self.yolo_model.output, self.anchors,
NameError: global name 'yolo_eval' is not defined

您好！我在使用deep_sort_yolov3，运行main时出现如上错误。咋办？

请问yolov3_sort里面是怎么确认检测的是行人的呢

hi,when i convert yolov3.weights to keras h5,the following error occurred.Please have a look. thanks!

python convert.py yolov3.cfg yolov3.weights model_data/yolo6.h5
Using TensorFlow backend.
Loading weights.
Weights Header: 0 2 0 [32013312]
Parsing Darknet config.
aaa <_io.BytesIO object at 0x7f5a9a4f5308>
Traceback (most recent call last):
File "convert.py", line 242, in
_main(parser.parse_args())
File "convert.py", line 81, in _main
cfg_parser.read_file(unique_config_file)
File "/opt/anaconda3/lib/python3.7/configparser.py", line 717, in read_file
self._read(f, source)
File "/opt/anaconda3/lib/python3.7/configparser.py", line 1030, in _read
if line.strip().startswith(prefix):
TypeError: startswith first arg must be bytes or a tuple of bytes, not str

如题 谢谢回复

I see that this function is implemented in SORT but not in Deep SORT

Feature: A feature vector that describes the object contained in this image. But what does this parameter represent? What is the meaning of the numbers in the array Feature?
For example: feature= [ 0.13135917 0.16252236 -0.02983991 0.1078044 0.18693653 0.222241
-0.07257525 -0.07441376 0.08326894 0.0541355 0.0026187 -0.07955563
-0.05495477 -0.05914702 -0.12457883 -0.04439378 -0.09501398 -0.12134293
0.02070528 -0.10250711 -0.02945639 -0.06187392 -0.11262443 -0.1067707
-0.12805346 -0.03683051 -0.01223208 -0.04542194 0.04128917 -0.09476566
0.13891572 -0.03608855 -0.08671472 0.17196652 -0.11042515 -0.09581856
-0.04645975 0.08574935 -0.09971982 -0.1007655 -0.11324123 0.09481143
-0.13796814 0.02062852 -0.03129434 0.07220582 -0.06094034 0.0997282
0.09703507 -0.11823436 -0.13719977 -0.1204574 0.0124543 -0.09292966
-0.03511186 -0.11681295 -0.06900121 -0.12038399 0.03251497 0.03251189
-0.0476766 -0.08023439 -0.00856623 0.06366589 -0.01062125 -0.07592105
-0.11542226 -0.01129758 -0.09487565 0.1098093 -0.03944121 0.15487668
0.06516567 -0.1064471 -0.11643744 -0.02732632 -0.11259694 0.00147146
0.04236715 -0.05228471 -0.07022281 -0.01324428 -0.10091629 -0.10923474
-0.04395907 -0.09888252 -0.08017209 -0.08089618 0.00113982 0.21809936
-0.08897579 -0.112206 0.0176195 -0.11458117 -0.00677603 0.0769835
-0.08030172 -0.01627187 0.02889497 -0.0008515 0.06290624 0.01271521
0.09568803 0.01536036 0.05172337 -0.09787396 -0.10839746 -0.069438
0.06256402 -0.0726283 -0.0296073 -0.10417195 -0.0954337 -0.09770833
0.0043749 -0.11574724 -0.00393125 0.05078177 -0.08470584 -0.12741965
-0.11512636 -0.0954428 0.01246825 0.06667607 -0.01297721 -0.14558725
-0.01562111 0.00888595]

Traceback (most recent call last):
File "main.py", line 172, in
main(YOLO())
File "main.py", line 158, in main
if len(pts[track.track_id]) != None:
UnboundLocalError: local variable 'track' referenced before assignment

Why would this happened?
Using deep sort yolov3

Hi,
1- Is it possible to track more than one classes at the same time? for example my model is trained on 6 classes how i can track all classes at the same time?

I am able to track one class successfully but unable to track all.
I have tried this method
if predicted_class != 'person' and predicted_class != 'car':
print(predicted_class)
continue

but it's tracking same by default class.

2-If we track more than one classes , will it effect the speed?

Hello @yehengchen, I am running YOLOv3 + Deep_SORT example, but I can't find the files required to run this example, how do I create them or can you provide them? That'd be great help.

deep_sort_yolov3/counter.txt
deep_sort_yolov3/num.txt
model_data/market1501.pb

训练 yolov3 spp net 网络命令？

What is the meaning of the numbers in the array feature? What does that represent?

hi
請問YOLOv3 + Deep_SORT裡readme寫說
2. Download [yolov3.weights] and place it in yolov3_sort/yolo-obj/
但run時缺少model_data/yolo_cc_0612.h5

請問這邊的model_data/yolo_cc_0612.h5是否是直接Download [yolov3.weights] and place it in deep_sort_yolov3/model_data/還是是您另外再train過weight？

运行时的错误提示：
tensorflow.python.framework.errors_impl.InvalidArgumentError: 2 root error(s) found.
(0) Invalid argument: Input to reshape is a tensor with 43095 values, but the requested shape requires a multiple of 5577
[[{{node Reshape_3}}]]
[[concat_12/_2923]]
(1) Invalid argument: Input to reshape is a tensor with 43095 values, but the requested shape requires a multiple of 5577
[[{{node Reshape_3}}]]
之前以为是视频的问题，调成640*480依然没有解决
代码定位应该是在yolo.py中我这边116行 out_boxes,out _scores这行开始

hi
我的環境是
rtx2070, cuda10, cudnn7.4, tensorflow-gpu 1.14
但無法work
會出現
Failed precondition: Attempting to use uninitialized value batch_normalization_61/beta
想請問你的執行環境，或者有沒有人在rtx或tensorflow-gpu 1.14上執行過。
.h5 及 .pb均有重新convert過。

How to calculate the respective time in Yolo+deepsort project?

呃...

if intersect(p0, p1, line[0], line[1]) and LABELS[classID[i]]=='car':
counter += 1

我在这里做了更改,希望只计数车辆,但是

text = "{}: {:.4f}".format(LABELS[classIDs[i]], confidences[i])
cv2.putText(frame, text, (x, y - 5), cv2.FONT_HERSHEY_SIMPLEX, 0.5, color, 2)

这样的话,测试我的视频,会发现将骑车的person识别为car甚至是traffic light- -

请问这是什么原因啊

As mentioned above, I would like to use my own training set. Can you provide training code for detection and tracking?

您好:

我在win10底下，使用python main.py --input input/test.mp4 --output output/test.avi --yolo yolo-obj
出現以下訊息
[INFO] loading YOLO from disk...
[INFO] 1283 total frames in video
[INFO] single frame took 1.0242 seconds
[INFO] estimated total time to finish: 1314.0222

CPU會接近100%，但是不會跳出影像的視窗，GPU也沒有工作

請問您知道這可能是什麼問題嗎?

Thanks for your sharing. Is there some project about real time detection on high resolution video ?