Thanh Tin Nguyen
Thanh Tin Nguyen
I have also run into this problem. I temporarily fixed it by commenting all related to the error. But when I run the /bin/imagefeaturedetector, I have an error: Segmentation fault...
I have fixed the issue by changing main.cpp a little bit: ` QApplication app(argc, argv); WindowMain mainWindow; mainWindow.show(); // new WindowMain(); return app.exec(); ` Hope it help !
Maybe qt5 on windows lacks some package. Actually, I have run on Ubuntu 18.04.
@ouafi98 I have source it in bashrc,for the 1) errror, when I run the launch_evaluation.bash, it can open like this, but the GUI for controling the drone is not working:...
@yun-long i havent solved the yaml file error. I use python3.6, the gui does not have any connect button, it only has off red button like the above image?
@gorghino yes I tried but nothing happened
@yun-long yes I sourced exactly my FLIGHTMARE_PATH and solved the YAML problem. I set up my workspace to work with python3 as well
When I click on the GUI the cmd output like this: 
I have a question that did you use your model to train on a custom data ? I see it is very difficult to make loss decrease
> I wonder if batch size has a great impact on the experimental results? I used two 3090ti to train the model (which means that the batch size is 4)....