Terrain Recognition using Transfer Learning

[DarshAgrawal14]

hey @Nancyjikadra
I am Darsh Agrawal , a fellow contributor and I would like to contribute by using Transfer learning and create a model to help various agencies label captured terrain images like mountains, coast , desserts etc . By using transfer learning we can use pre trained models like RESNET or Efficientnet and hyperparameter tune them for terrain classification for better accuracy when compared to traditional CNN models.

[github-actions]

Thanks for creating the issue,Please read the Pinned issued first and Readme.md in each Pull Request you made. Keep learning...

[Nancyjikadra]

Hi @DarshAgrawal14 , The idea is good. Just to confirm, are you asking if I’d like to team up with you on this project? Let me know!

[DarshAgrawal14]

Hi @Nancyjikadra,
I just wanted to clarify that I'm offering to train the model using transfer learning and then submit a PR. If you're interested, I can proceed with that and send over the results for you to review.
Thanks!

[Nancyjikadra]

Yup, sounds good. you can proceed, looking forward to seeing the results. Happy coding! @DarshAgrawal14

[DarshAgrawal14]

hello @Nancyjikadra , can you please rename this folder "Quality Education SDG 04: Predictive Analysis" because this folder naming is creating path error while cloning repo , even if ignored while cloning the files will get deleted while submitting PR as they were not in the local repo :

[DarshAgrawal14]

once the renaming is done , i think i should be able to submit PR

[DarshAgrawal14]

here is a video to show working of the model :

Untitled.video.-.Made.with.Clipchamp.5.mp4