Bootstrapping User-Defined Body Tapping Recognition with Offline-Learned Probabilistic ...
summary
Combines an offline-learned probabilistic representation of body taps with a few user-provided templates on commodity phone motion sensors and proximity-triggered tap detection, then classifies new taps by KL divergence matching to those probabilistic templates. Claims a hybrid system that needs only a single sample per location to improve accuracy by ~30.8% over an SVM baseline and ~24.8% over template matching in user studies, supports unconstrained body locations and natural tapping rates.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 0:14Background
- 1:50Current Approach
- 4:38Experiment
- 5:26Simulation
- 6:30Analysis
- 8:28Summary
related talks
05mFine-Grained and Context-Aware Behavioral Biometrics for Pattern Lock on Smartphones
19mATK: Enabling Ten-Finger Freehand Typing in Air Based on 3D Hand Tracking Data
27mDissecting ActivityRecognition with Multi-Sensor Simulation and Its Practical Deployment
10mHandyTrak: Recognizing the Holding Hand on a Commodity Smartphone from Body Silhouette Images
06mRobust Unsupervised Factory Activity Recognition with Body-worn Accelerometer Using Temporal ...
28mDeep Learning-Based Human Activity Recognition for Physical Skills...
06mSwingNet: Ubiquitous Fine-Grained Swing Tracking Framework via Stochastic Neural Architecture ...
19mGroup Touch: Distinguishing Tabletop Users in Group Settings via Statistical Modeling of Touch Pairs