ViscoCam: Smartphone-based Drink Viscosity Control Assistant for Dysphagia Patients
summary
A smartphone app that measures drink viscosity by having the user shake a cup and phone while the flashlight camera images refraction distortions and the built‑in accelerometer or microphone records motion, then computes optical flow and feeds it to a 3D convolutional neural network to output viscosity. It classifies three IDDSI levels, achieves 96.52% accuracy in controlled tests and stays >81% under extreme cases, robust to cup size, liquid height and color.
AI-written summary · about
chapters (9)
- 0:00Intro
- 0:13Dysphagia
- 0:39About Diet Modification
- 1:31Drink Viscosity Control: Current Methods
- 2:59ViscoCam
- 3:19Liquid Sloshing
- 4:04Sensing the Liquid Surface Wave Height
- 5:03System Overview
- 5:28Evaluation
related talks
21mVizLens: A Robust and Interactive Screen Reader for Interfaces in the Real World
18mCamTrackPoint: Camera-Based Pointing Stick Using Transmitted Light through Finger
06mInSight: Monitoring the state of the driver in low-light using smartphones
21mA Face Recognition Application for People with Visual Impairments: Understanding Use Beyond the Lab
41mComputer Vision for Remote AR
19mSymmetriSense: Enabling Near-Surface Interactivity on Glossy Surfaces using a Single Commodity ...
15mMeCap: Whole-Body Digitization for Low-Cost VR/AR Headsets
06mProviding Better Feedback in Real-time Object Detection Apps