-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Tim Weißker edited this page Aug 3, 2016
·
16 revisions
Welcome to the Smartwatch Handshake (SoSe 2016) wiki!
##Recorded smartwatch data
- An initial archive with some recorded single- and double-sided handshakes (gravity excluded) as well as null gestures can be downloaded at www.uni-weimar.de/~lopa1693/handshake/smartwatch-data.zip
- Andi and Tim shake hands with gravity included: www.uni-weimar.de/~lopa1693/handshake/andi-tim-gravity-shakes.zip
- Some demo (non-)handshakes with the new gravity trigger mechanism: www.uni-weimar.de/~lopa1693/handshake/triggered-training.zip
- Erdan and Tim shake hands with gravity included: www.uni-weimar.de/~lopa1693/handshake/erdan-tim-gravity-shakes.zip
- Andi and Frederik shakes data l and r for shake with 2 watches. https://drive.google.com/open?id=0BwokxLtfisRdWFlvQlFETF8zUHc
##Related work
- Recognizing Text Using Motion Data From a Smartwatch
- recognize text written on a whiteboard using motion data from a smartwatch
- http://www.tik.ee.ethz.ch/file/cd34078b1f05c851e57f1522fb61961f/wristsense.pdf
- Personalizing a smartwatch-based gesture interface with transfer learning
- interface for visually impaired people allowing to train custom gestures for interaction
- uses novel machine learning algorithm called transfer learning from known and trained gestures
- http://www.eurasip.org/Proceedings/Eusipco/Eusipco2014/HTML/papers/1569922319.pdf
- WristSnoop - Smartphone PINs prediction using smartwatch motion sensors
- The Office Smart Watch
- reacts to gestures in the office context, e.g. virtual door knocks or (un)locking a door
- https://uhdspace.uhasselt.be/dspace/bitstream/1942/16940/1/BernaertsDruweSteenselsVermeulenSchoning_dis-pwip2014.pdf
- MoLe: Motion Leaks through Smartwatch Sensors
- can recognize text the user is typing using smartwatch sensors
- http://www.sigmobile.org/mobicom/2015/papers/p155-wangA.pdf
- Finger-writing with Smartwatch: A Case for Finger and Hand Gesture Recognition using Smartwatch
- recognizes 37 different arm, hand and even finger gestures
- http://dl.acm.org/citation.cfm?id=2699350
- The Gesture Recognition Toolkit
- makes machine learning and gesture recognition more accessible for non-specialists
- free C++ library, can be ported to the Android NDK
- http://jmlr.org/papers/volume15/gillian14a/gillian14a.pdf
- GitHub: https://github.com/nickgillian/grt
- Introductory Talk Slides: http://www.nickgillian.com/archive/teaching/workshops/mitiap2013/IAP_Session1.pdf
- Apple patent 25.06.2015:
- https://patentscope.wipo.int/search/en/detail.jsf?docId=WO2015094220&recNum=53&maxRec=51457&office=&prevFilter=%26fq%3DOF%3AWO%26fq%3DICF_M%3A%22H04W%22&sortOption=Data+pub+ordem+inversa&queryString=&tab=PCTDescription
- http://worldwide.espacenet.com/publicationDetails/originalDocument;jsessionid=Gwzsvx6InqCQeUxMsnbzHD52.espacenet_levelx_prod_3?CC=WO&NR=2015094220A1&KC=A1&FT=D&ND=&date=20150625&DB=&&locale=en_EP
- Multi-Dimensional Dynamic Time Warping for Gesture Recognition
- explanation of the DTW algorithm
- http://msp.ewi.tudelft.nl/sites/default/files/DTW-vASCI.pdf
- Accelerometer-Based Hand Gesture Recognition using Feature Weighted Naïve Bayesian Classifiers and Dynamic Time Warping
- Angel Sensor
- Angel Sensor is the only wearable designed as an open platform for mobile health. It tracks heart rate, skin temperature, steps, sleep quality, calories, acceleration, and orientation. It offers unrestricted, real-time API to its sensors and full ownership of the data for both developers and consumers
- http://angelsensor.com/
- RetroBand - Make your own activity tracker
- based on Arduino
- http://www.instructables.com/id/Make-your-own-activity-tracker/
- Automatic standalone Contact Information Exchanger:
- GitHub Handshake app:
- konzept:
- Smartbands:
- Smartband 2
- UP2
- Fibit Flex
- Underamor Band
- Angle Sensor
- Microsoft Band 2
- Misfit Flash
- Xiami Mi -> https://github.com/Freeyourgadget/Gadgetbridge