I’ve trained YOLOv9 and YOLOv10 on original data for object detection and created a GUI application to demonstrate their capabilities. Please check out the GitHub repository in the description and try it out!”
Thanks for using the code! To use the video detection feature, select the "Image/Video" tab in the sidebar and choose the folder where your video is saved. It will process the video and save the results in the "results" folder. The main window of the GUI only displays images after detection, which is by design.
@@spreadknowledge Thank you it can be run with CUDA(GPU)? your doc don written about it. In your video, the TRAIN PART is confusing. In the video, the IMAGE folder and labeling folder are separated, so it appears as if the progress is being made even if the .img file and .txt file are not in the same folder. However, in reality, files of both formats must be in one space to proceed. I also want to say thank you to you. It is much more effective to teach yolo to juniors through a GUI than to teach it through statement code. Among yolo GUIs, there are not many TRAIN and DETECTION options together. I feel grateful to you.
@@spreadknowledge If torch version 2.3.0 is CUDA 1.2.10, does the Train result not appear? As a result of running it in the current state, there is nothing in the weight folder in the model save folder I specified, and the train data has also disappeared.
@@user-rm7zu7py9z Thank you for your questions and feedback! 1. You can run it with CUDA. Please ensure you have the correct version of PyTorch that matches your CUDA version. 2. Thank you for pointing out the confusion regarding the training part in the video. Initially, all image and txt files should be in the same folder. They will automatically be split into train:val=8:2. If you have already manually split them into train and val, with separate images and labels subfolders, the program will proceed without splitting. 3. I appreciate your report about the bug in saving the training results. I have debugged cases with 100 epochs, and it might not work correctly for fewer epochs. I will fix this bug soon. 4. Thank you for your kind words about the program. I also wanted a platform that could perform both model training and object detection. I hope more people realize the convenience of this app, and I will incorporate your feedback into the code to improve it further.
@@spreadknowledge Thank you for your reply. The error occurred because i set the epoch to 50! If I have any questions in the future, can I contact you via email or reply? For reference, there is a problem in the second answer where even though it is divided, it is not recognized and cannot be processed.
I’ve trained YOLOv9 and YOLOv10 on original data for object detection and created a GUI application to demonstrate their capabilities.
Please check out the GitHub repository in the description and try it out!”
is this can use video ?? in your git introduced with video but i attempt it is wrong
Thanks for using the code!
To use the video detection feature, select the "Image/Video" tab in the sidebar and choose the folder where your video is saved.
It will process the video and save the results in the "results" folder.
The main window of the GUI only displays images after detection, which is by design.
@@spreadknowledge Thank you it can be run with CUDA(GPU)? your doc don written about it.
In your video, the TRAIN PART is confusing. In the video, the IMAGE folder and labeling folder are separated, so it appears as if the progress is being made even if the .img file and .txt file are not in the same folder. However, in reality, files of both formats must be in one space to proceed.
I also want to say thank you to you. It is much more effective to teach yolo to juniors through a GUI than to teach it through statement code. Among yolo GUIs, there are not many TRAIN and DETECTION options together. I feel grateful to you.
@@spreadknowledge
If torch version 2.3.0 is CUDA 1.2.10, does the Train result not appear? As a result of running it in the current state, there is nothing in the weight folder in the model save folder I specified, and the train data has also disappeared.
@@user-rm7zu7py9z Thank you for your questions and feedback!
1. You can run it with CUDA. Please ensure you have the correct version of PyTorch that matches your CUDA version.
2. Thank you for pointing out the confusion regarding the training part in the video. Initially, all image and txt files should be in the same folder. They will automatically be split into train:val=8:2. If you have already manually split them into train and val, with separate images and labels subfolders, the program will proceed without splitting.
3. I appreciate your report about the bug in saving the training results. I have debugged cases with 100 epochs, and it might not work correctly for fewer epochs. I will fix this bug soon.
4. Thank you for your kind words about the program. I also wanted a platform that could perform both model training and object detection. I hope more people realize the convenience of this app, and I will incorporate your feedback into the code to improve it further.
@@spreadknowledge Thank you for your reply. The error occurred because i set the epoch to 50! If I have any questions in the future, can I contact you via email or reply?
For reference, there is a problem in the second answer where even though it is divided, it is not recognized and cannot be processed.