Member Avatar for killer88

I am learning Android using java. i want to make a mini project on that.i wanna make an app which recognise face with name in it. can i make it ?

thanks

  • 3 Contributors
  • 3 Replies
  • 157 Views
  • 17 Hours Discussion Span
  • Latest Post Latest Post by azareth

Dani AI

Generated

Yes — you can build a face-recognition app in Android with Java. Split the job into two clear parts: face detection (locate faces in camera frames) and face recognition (map a detected face to a stored name). gave the right tip about using libraries; that will save a lot of time. also pointed you toward learning Android fundamentals first — that is important before tackling camera and model work.

A practical, beginner-friendly path:

  • Get camera input and permissions working (use CameraX). See the official CameraX guide: Android CameraX.
  • Add face detection first (easy): use ML Kit face detection to get bounding boxes and landmarks: ML Kit Face Detection.
  • For recognition (names) pick one approach:
    • Cloud APIs (fast to prototype): Azure Face or AWS Rekognition — they return identities if you register faces.
    • On-device classical method: OpenCV LBPH (simpler, less accurate): OpenCV.
    • On-device deep method: extract embeddings with a pretrained model and compare with TensorFlow Lite: TensorFlow Lite.

Typical pipeline (high level):

Camera -> ML Kit detector -> crop + align face -> TFLite Face model -> embedding[] -> nearest neighbor -> show name

Practical tips and pitfalls:

  • Detection != recognition. You must collect labeled images per person and store embeddings or templates.
  • Preprocess: correct rotation, align eyes, resize to model input, normalize pixels.
  • Compare embeddings with cosine similarity and set a threshold for “unknown”.
  • For realtime use: run inference on a background thread, quantize the model, or use GPU delegate for speed.
  • Mind privacy and permissions: inform users and secure stored face data.

Start with the CameraX + ML Kit quickstarts, then try a TFLite embedding example or OpenCV sample depending on which route you choose. The links above point to stable official docs and sample material to follow next.

Recommended Answers

All 3 Replies

Member Avatar for rubberman

Sure. How good are you at image processing? Or, how good are your Google skills to find an already-written library to do that? :-) Myself, I could write it, but why? Leveraging the work of others who really know this stuff and who release it into the wild (open source projects) is not a "bad thing", unless you are doing it for school work! Then, it is cheating! :-)

Member Avatar for killer88

just i had an idea man of making an app for face recognisation. i never knew it has been already built. As i am a beginner ,Can i get some tutorial links for learining android?

thanks man :)

Member Avatar for azareth

Can i get some tutorial links for learining android?

you have google.. but anyway try this
http://www.coreservlets.com/android-tutorial/
it has some zip files that you can download for a better understanding in android programming

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.