ML-Kit is a free-to-use machine learning API for Android and iOS apps, focusing on on-device AI solutions.
In this repo, I have made a demo of all the AI techniques I have performed with ML-Kit on Android.

The ML techniques used in this demo app include but are not limited to:

1. Face mesh detection

This algorithm detects face mesh info on close-range images.

Example results:



2. Text recognition

This algorithm can recognize and extract text from images.

Some examples:



3. Pose detection

Detecting the position of human body in any given picture/video in real-time.

Example result:



4. Selfie segmentation

Separates the background of a picture/video from users within it. Helps to focus on more important objects in the picture/video.





5. Object detection

Localize and tag in realtime one or more objects in the live camera feed.

Example results of object detection:



6. Barcode scanner

Scanning and processing most kinds of barcodes. Supports various standard 1D and 2D (a.k.a. QR) barcode formats.

Example result of barcode scanner:



A video demo of barcode scanner:

Realtime barcode scanner video demo
barcode_demo.mp4

7. Image labeling

This algorithm identifies objects, locations, activities, animal species, products and more in a given picture.
For example, in the picture below, it has managed to label the road, Jeans, Jacket, and Buildings in the picture correctly.



8. Face detection

Detects faces and facial landmarks in a given image/video.

Example results:



9. Live Camera translator

This amazing demo camera app firstly determines the language of a string of text with just a few words. And then, translates that text between 58 languages; completely on device.

A video demo of realtime camera translator:

Live camera translator video demo
live.camera.translator.mp4

10. Digital ink recognition

This part of the app recognizes handwritten text and hand-drawn shapes (such as emojis) on a digital surface, such as a touch screen.

Some examples:



A video demo of digital ink recognition:

Digital ink recognition video demo
2023-04-11.01.05.42.mp4

Please star and fork this repo, I will be maintaining it over time and will try to add more ML related library demos to it.
Feel free to open issues and point out the bugs and short-comings.

new architecture of the app is going to be like this: we will have these 5 main modules in our codebase:

1. feature_ink_recognition
2. feature_live_translator
3. feature_still_image
4. feature_live_preview
5. shared