A mobile application to create accelerometer based activity recognition models directly on the phone. The configuration of the segmentation and feature extraction process chain requires expert knownledge. The prototype was developed in 2012 in a bachelor thesis at the University of Kassel and was optimized and enhanced for an experiment in 2015.

Features

  • accelerometer bases activity recognition
  • research project
  • individual training on the mobile phone
  • no data transmission for privacy protection
  • configurable inference process

Project Samples

Application components AM-Miner main screen AM-Miner activity sample recording screen AM-Miner crossvalidation on training data AM-Miner configuration screen - Part 1 AM-Miner configuration screen - Part 2

Project Activity

See All Activity >

Categories

Mobile, Machine Learning

License

GNU Library or Lesser General Public License version 3.0 (LGPLv3)

Follow android-activity-miner

android-activity-miner Web Site

You Might Also Like
Passwordless authentication enables a secure and frictionless experience for your users | Auth0 Icon
Passwordless authentication enables a secure and frictionless experience for your users | Auth0

Make your login flow seamless with Passwordless

Over two-thirds of people reuse passwords across sites, resulting in an increasingly insecure e-commerce ecosystem. Learn how passwordless can not only mitigate these issues but make the authentication experience delightful. Implement Auth0 in any application in just five minutes
Sign up
Rate This Project
Login To Rate This Project

User Reviews

Be the first to post a review of android-activity-miner!

Additional Project Details

Operating Systems

Android

Languages

English

Programming Language

Java

Related Categories

Java Mobile Software, Java Machine Learning Software

Registered

2015-09-01
Report inappropriate content