I tried creating my own custom dataset (my own 2 second videos and then had frames extracted in folders named with custom video id). Used ffmpeg to extract 12 frames per second. Used 10 videos per class, and 5 for validation (around 1000 images)
I then adapted csv files (train and validation ones) to add the required video ids for each of my custom classes - I tried this with 3 classes.
First issue is that code uses deprecated functions regarding Series.from_csv and failed. I had to convert csv to json files in order to continue.
Then I trained with 20 epochs, then tried test-app.py (after adapting my own class names) but nothing is predicted, no gesture although training results seems ok.
Would it be possible to have like a small guide in here on how to use your example with custom data (file/folder structure/sizes etc) and maybe revisit code so it works with latest libraries etc?
Hello,
I tried creating my own custom dataset (my own 2 second videos and then had frames extracted in folders named with custom video id). Used ffmpeg to extract 12 frames per second. Used 10 videos per class, and 5 for validation (around 1000 images) I then adapted csv files (train and validation ones) to add the required video ids for each of my custom classes - I tried this with 3 classes.
First issue is that code uses deprecated functions regarding Series.from_csv and failed. I had to convert csv to json files in order to continue. Then I trained with 20 epochs, then tried test-app.py (after adapting my own class names) but nothing is predicted, no gesture although training results seems ok.
Would it be possible to have like a small guide in here on how to use your example with custom data (file/folder structure/sizes etc) and maybe revisit code so it works with latest libraries etc?
Thank you very much