Skinput: Appropriating the Body as an Input Surface
Download as PPTX, PDF•1 like•410 views
Skinput is a technology that uses the human body as an input surface by sensing vibrations on the skin caused by taps and gestures. An array of vibration sensors can detect finger taps on different locations on the arms and hands with over 80% accuracy on average. Tests showed higher accuracy for taps below the elbow compared to above. Identification was slightly reduced for participants with higher BMI. Further experiments explored uses while walking/jogging and recognizing single-handed gestures, surfaces, finger tap types, and segmented finger input. Skinput provides an alternative to small device screens for mobile interaction.
Skinput: Appropriating the Body as an Input Surface
- 1. Skinput: Appropriating the Body as an Input Surface Chris Harrison, Desney Tan, Dan Morris Summarized & Presented by: Reem Alattas
- 2. Skinput • A technology that appropriates the human body for acoustic transmission, allowing the skin to be used as an input surface.
- 3. Introduction • Handheld devices – Size vs. Computational Power • Alternative interaction methods
- 4. Related Work • Always-Available Input – Skin – Computer Vision – Speech Input – Wearable Computing • Glove-based Input • Smart Fabric • SixthSense • Computationally expensive • Error prone in mobile scenarios • Limited precision • Privacy and scalability issues
- 5. Related Work • Bio-Sensing – Brain sensing technologies • EEG • fNIR
- 6. Skinput • An input technique that allows the skin to be used as a finger input surface.
- 7. Bio-Acoustics • Transverse wave propagation • Longitudinal wave propagation
- 8. Sensing • An array of highly tuned vibration sensors
- 9. Processing
- 10. Experiment • 13 Participants (7 females) • Seattle area • Age range: 20 – 56 (mean 38.3) • BMI range: 20.5 (normal) – 31.9 (obese) • 1 lefty
- 11. Experimental Conditions • Fingers (Five Locations) • Whole Arm (Five Locations) • Forearm (Ten Locations)
- 12. Procedure • Participants practiced tapping approximately one minute with each gesture set. • Participants were instructed to comfortably tap each location ten times, with a finger of their choosing. – 3 rounds of training data were collected per input location set (30 examples per location, 150 data points total) – Forearm Exception: two rounds were collected (20 examples per location, 200 data points total)
- 13. Results • Five Fingers – Average accuracy 87.7% • Whole Arm – Below elbow average accuracy 95.5% – Above elbow average accuracy 88.3% • Forearm – Average accuracy 81.5%
- 14. Accuracy
- 15. BMI Effects • High BMI is correlated with decreased accuracies.
- 16. Supplemental Experiments • Walking and Jogging • Single-Handed Gestures • Surface and Object Recognition • Identification of Finger Tap Type • Segmenting Finger Input
- 17. Conclusion • The proposed approach appropriates the human body as an input surface.
- 18. Questions
Editor's Notes
- #2: http://www.youtube.com/watch?v=eK9GdNq4fnc
- #4: Inverse relationship Since scientists cannot simply make buttons and screens larger without losing the primary benefit of small size, they consider alternative approaches that enhance interactions with small mobile systems. Using the surface that happens to always travel with us: our skin.
- #5: The primary goal of Skinput is to provide an always available mobile input system – that is, an input system that does not require a user to carry or pick up a device.
- #6: Skinput leverages the natural acoustic conduction properties of the human body to provide an input system, and is thus related to previous work in the use of biological signals for computer input.
- #7: They choose to focus on the arm as it provides considerable surface area for interaction, including a contiguous and flat area for projection. The forearm and hands contain a complex assemblage of bones that increases acoustic distinctiveness of different locations.
- #8: When a finger taps the skin, several distinct forms of acoustic energy are produced.
- #9: Generates a response curve – not flat. Figure 4 shows the response curve for one of our sensors, tuned to a resonant frequency of 78Hz. The curve shows a ~14dB drop-off ±20Hz away from the resonant frequency.
- #10: Ten channels of acoustic data generated by three finger taps on the forearm, followed by three taps on the wrist. The exponential average of the channels is shown in red. Segmented input windows are highlighted in green. Note how different sensing elements are actuated by the two locations.
- #18: approach to appropriating the human body as an input surface