Real Time Object Detection with YOLOv9 and Webcam

[@aaryadesai6224]

I faced the same issue of no module named torch. In the directory of YOLOV9 there is a file called requirements.txt, this file lists all the required dependencies. use this command in the terminal (pip install -r requirements.txt). this installed everything for me. I faced another issue where it could not find the source. I am using a usb webcam, but had to change my source from 1 to 0 for it to work. Thanks for this amazing and easy tutorial.

[@tech_watt]

Love it 😍 thanks for your explanation. You’ve help me a lot 👍

[@GamerDudeBYT]

Thank you for this video. You made it very straightforward and very useful. I have been looking for 5 months for a good object detection video and you have done that.

[@tech_watt]

Thanks 😊 so much for your kind words. Am super excited you’ve finally found this tutorial

[@aiforyounow]

Great tutorial dude. You make everything very easy to understand.

[@Deepsim]

Very great tutorial on yolov9. Brother, how to print the bbox and center of it? Thanks

[@tech_watt]

Tutorial on that coming soon

[@kristijanlazarev]

It's so awesome and simple damn!!

[@tech_watt]

Thanks 😊

[@EiNETCORPPvtLtd]

Great video!! dude. Can you please make a video using CSI camera

[@tech_watt]

Yes but I don’t have it

[@dakaradakaradakara699]

Please do this with the custom dataset on roboflow ?

[@samuraix2193]

I faced some issue it say 'Phyton was not found; run without arguments to install from the Microsoft store, or disable this shortcut from Settings > Manage app execution aliases. ' can you help me

[@BangtanArmy-mt7ph]

Import torch is not resolved

[@YigalBZ]

Thank you for the good video. Very easy to follow and understand! If my laptop has a P1000 GPU of Nvidia, how to use it? Just by writing gpu instead of cpu?

[@tech_watt]

In that case you just write 0. But if the gpu is not configured, you will get error

[@YigalBZ]

@@tech_watt I am not sure I understand - configure where? It exists there, just like the CPU.

[@tech_watt]

@YigalBZ no I mean you have to set up your GPU on your computer before it will be able to pick it.

[@GamerDudeBYT]

Do you know how to get it work with a video? I tried just making the source with a video but it never stops the script and the video is not playable

[@tech_watt]

When you give source to a video file it will detect objects in the video file and when it’s done running it will save it in a folder 📂 path called runs. The name of the path will be written on the terminal.

[@GamerDudeBYT]

@@tech_watt Thanks. I had tried that but the video was not working but i found that adding --view-img on the end of the command showed it in real time

[@Rapper_Ash]

bro it is showing that no torch module is showing what to do for that

[@tech_watt]

Before running the code make sure you’re in same directory you’ve clone the repository and also do install all the dependencies in the requirements file

[@varshhadinesh7538]

hello!! thank you for the great video. I just wanted to ask whether this project works only for torch version 1.7.0 or will it work for any other latest version? I'm facing an error while running the project as I couldn't able to download torch. error: Traceback (most recent call last): File "C:\Users\varsh\OneDrive\Desktop\yolo v9 det\yolov9\detect.py", line 231, in main(opt) File "C:\Users\varsh\OneDrive\Desktop\yolo v9 det\yolov9\detect.py", line 226, in main run(**vars(opt)) File "C:\Users\varsh\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context return func(*args, **kwargs) File "C:\Users\varsh\OneDrive\Desktop\yolo v9 det\yolov9\detect.py", line 89, in run im = torch.from_numpy(im).to(model.device) NameError: name 'torch' is not defined please suggest me a solution for this. TIA

[@tech_watt]

You can pip install all the packages in the requirements. txt file to get all the packages installed

[@varshhadinesh7538]

@@tech_watt can you please tell me the command? I've already tried pip install -r requirements.txt

[@tech_watt]

@varshhadinesh7538 yes that’s the command but make sure you are in the directory before executing any command

[@tech_watt]

@varshhadinesh7538 just follow all the steps in the tutorial

[@varshhadinesh7538]

@@tech_watt Thank youu!! got the output :)

[@Shashika_Janith]

Thank you

[@tech_watt]

You're welcome

[@jaimeabug9090]

it said Traceback (most recent call last): File "detect.py", line 7, in import torch ModuleNotFoundError: No module named 'torch'

[@tech_watt]

Make sure you install all the dependencies in the requirements.txt file

[@jaimeabug9090]

@@tech_watt where should I see them all?

[@kristijanlazarev]

No module named torch

[@tech_watt]

Kindly do pip install requirements.txt to install all the necessary packages from the clone repository

[@kristijanlazarev]

@@tech_watt But it still doesn't work for some reason, hmmm

[@kristijanlazarev]

Nvm it got fixed because i was using py instead of python in the command. Isn't it same?

[@tech_watt]

@kristijanlazarev that’s great 😀. They are not the same py is just the extension

[@mecha.mio.00]

@@tech_watt i have the same problem with the module "torch". in requirements txt, i just put avaible the lines 1 and 2 of pipi install requeriements.txt, but it still doesn t work. do you have an idea what is happend? PS C:\Users\MIO\Desktop\yolov9> python detect.py --weights gelan-c.pt --conf 0.5 --source 1 --device 0 Traceback (most recent call last): File "C:\Users\MIO\Desktop\yolov9\detect.py", line 7, in import torch ModuleNotFoundError: No module named 'torch' PS C:\Users\MIO\Desktop\yolov9>