arunkpatra / bow-wow Goto Github PK

Would you like to help create Bow-Wow?

Please comment on this issue, if you would like to contribute. We would need iOS and Android developers, data scientists, vets and programmers. Actually, you know what? if you are none of the above and just love your pet, it's probably just enough. We can learn everything on the way...

The idea is to work with a real time video stream and generate sentiment information on the fly. We would then superimpose live sentiment information on-screen. The Android app can be developed in isolation from the algorithms that would provide sentiment data.

Consider native libraries of the iOS and Android platforms first before considering OpenCV's camera abstraction. Use whatever tool can give the most stable, highest performance, lowest resource usage and most pleasant programming experience(even at the cost of higher development effort). If we are not happy with existing libraries, is it worth writing our own? I think we should not be afraid of that.

The idea is to work with a real time video stream and generate sentiment information on the fly. We would then superimpose live sentiment information on-screen. The iOS app can be developed in isolation from the algorithms that would provide sentiment data.

• Gather dog image datasets
• How do we tag data
• What sentiments we choose for now
• Are vets needed to provide expert guidance now?
• Should dog trainers be involved.
• Get as much help as needed.

By the end of this task we should have everything to get started with designing the app.

This is a big one, and probably will get split into multiple issues in the future.

The objective would be, to create the right models which can be consumed in the constraints environment of a mobile device. We can do the training in whatever environment we can get our hands on. Anyone with spare compute power, are most welcome to help with the training.

Develop the API layer that will expose operations of the sentiment calculation algorithms. Consider Kotlin for Android and Swift for iOS.