Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fine-tuning #4

Open
fiastros opened this issue May 5, 2021 · 2 comments
Open

Fine-tuning #4

fiastros opened this issue May 5, 2021 · 2 comments
Labels
question Further information is requested

Comments

@fiastros
Copy link

fiastros commented May 5, 2021

I'm working on cat facial recognition. so my first idea was to calculate the distance between two embedded cat faces.
Is it possible to retrain/fine tune the embedding models for such task ?

Thank you.

@fiastros fiastros added documentation Improvements or additions to documentation good first issue Good for newcomers help wanted Extra attention is needed labels May 5, 2021
@oke-aditya oke-aditya added question Further information is requested and removed documentation Improvements or additions to documentation good first issue Good for newcomers help wanted Extra attention is needed labels May 5, 2021
@oke-aditya
Copy link
Owner

Yes of coure you can retrain these models on any dataset. The training scripts are provided here.
Note that this similarity search works on unlabelled data. It can help to compute distance between two cat faces.

But a few thoughts about cat facial recongition.
You can probably label the data and train a object detector such as FasterRCNN. These are available in torchvision. It would work better and faster.

@fiastros
Copy link
Author

fiastros commented May 5, 2021

Thank you for your advice.
I have tried yolov5 for object detection: it works fine but the issue is that it needs lots of images for new faces, and it needs retrain each time....

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
question Further information is requested
Projects
None yet
Development

No branches or pull requests

2 participants