Pages

Class 17, Reading 1

Reading

  • GestureWrist and GesturePad: Unobtrusive Wearable Interaction Devices
  • Rekimoto

Outline

  • 1. Introduction
  • 2. Related work
  • 3. Technological background
    • 3.1. Capacitance sensing
    • 3.2. On-body networking
  • 4. GestureWrist A wristband-type input device
    • 4.1. Hand-gesture recognition
    • 4.2. Forearm movement measurement
    • 4.3. Tactile feedback
    • 4.4. Combining two sensor inputs
  • 5. GesturePad: A sensor module for interactive clothing
    • 5.1. Sensor configurations
    • 5.2. Interaction areas on clothing
    • 5.3. Clothes as communication surfaces
  • 6. Discussion and future research
    • 6.1. Combination with output devices
    • 6.2. Combination with on-body networks
    • 6.3. Sensing other parts of the body

Notes

  • Introduction
    • Important wearable computing feature #1: Supporting hands-free operations or allowing quick changes between normal and operation modes
      • Some input devices used with wearable computers need to be held during operation
      • With one hand already occupied, performing two-handed tasks becomes difficult
      • Glove-type input devices hinder performing real-world tasks
    • Important wearable computing feature #2: Social acceptance
      • Input devices should be as natural and unnoticeable as possible for use in various social settings
      • Many current wearable input devices look unusual or too "hi-tech"
        • Often troublesome to wear long-term in everyday situations
      • Traditional wearable accessories (e.g., wristwatches, eyeglasses, jewelry, clothes) part of daily life and designed for comfort with long-time use
    • "Unobtrusiveness"
      • Unobtrusiveness of input devices essential for being used in everyday situations
      • One possible way is to embed input sensors to conventional wearable items
    • Paper's two example devices:
      • GestureWrite
        • Wristwatch-type input device that recognizes human hand gestures by capacitively measuring changes in wrist type
        • Combined with acceleration sensor mounted to wristband, can be used as command-input device
      • GesturePad
        • Layer of sensor electrodes transforms conventional clothes into interaction devices (i.e., "interactive clothing")
        • Can be attached to area of clothing such as sleeve or lapel
        • On capacitive sensing, can detect and read finger motions applied on outside of clothing fabric
  • Technological background
    • Capacitance sensing
      • Definition:
        • A technique for measuring distances of nearby conductive objects
        • Measures the capacitance between the sensor and the object and uses a transmitter and a receiver electrode
      • Used for measuring:
        • the arm shape by placing both the transmitter and receiver electrodes on a wristband
        • finger positions by attaching electrodes on the inside of clothes
    • On-body networking
      • Based on capacitive sensing, a technique that transmits data through the human body
      • Wearable devices can:
        • communicate with each other
        • automatically authenticate digital devices that are touched
      • Technique used to distinguish wearer from other people while interacting with GesturePad
  • GestureWrist
    • About
      • A wristwatch-type input device that recognizes human hand gestures
      • Capacitively measures:
        • wrist-shape changes
        • forearm movements
      • Consists of:
        • two input sensors (capacitance and acceleration sensors)
        • one tactile feedback actuator
      • fabricated by attaching sensors and actuators to wristwatch
      • embedding all sensing elements into wristwatch technically possible, so can be used in any social situation
    • Hand-gesture recognition
      • Recognizers hand gestures by measuring changes of arm shape on inside of wristband
      • Handled by combination of transmitter and receiver electrodes attached to back of watch dial and inside wristband
      • When wearer opens and closes hand, wrist's cross-sectional shape changes accordingly
      • System can clearly distinguish two hand shapes:
        • Grasping
        • Pointing
    • Forearm movement measurement
      • Uses solid-state 2-axis acceleration sensor
      • Measures inclination of forearm
    • Tactile feedback
      • When gesture is recognized, GestureWrist gives tactile feedback to user
      • Actuator inside wristwatch dial provides feedback
    • Combining two sensor inputs
      • By combining two inputs, simple gesture commands were designed
      • Two hand shapes:
        • Fist
        • Pointing
      • Six arm positions:
        • Palm up
        • Palm right
        • Palm left
        • Palm down
        • Forearm up
        • Forearm down
      • Gesture commands
        • Hand shapes used to separate gesture commands into segments
        • Two consecutive arm positions (e.g., palm left -> palm down) make up one input command
  • GesturePad: A sensor module for interactive clothing
    • About
      • Trial: transform conventional clothes into interactive objects
      • Previous work on interactive clothes
        • Used metallic yarns woven into fabrics
        • Difficult to apply to clothes that already exist
      • "Retrofit" approach
        • Allows users to attach interactive modules on clothes easily
        • Concentrated on making attachment as unnoticeable as possible
        • Clothes are highly social media, so attaching obtrusive devices not ideal solution
      • GesturePad consists of a layer of sensors that can be attached to inside of clothing
        • Wearer can control module from the outside
        • Part of clothing becomes interactive without changing appearance
    • Sensor configurations
      • Consists of three configurations
        • Can be attached to inside of clothes
        • Controlled from the outside
      • Configuration #1: Type-A
        • Consists of an array of capacitive sensors
        • When user's finger is close enough to sensor surface, sensor grid recognizes finger position
        • E.g., when module is placed on inside of lapel, finger stroke gesture on lapel becomes computer input
      • Configuration #2: Type-B
        • Consists of transmitter and receiver layer separated by shield layer
        • When user's finger is within proximity of GesturePad, wave signal from transmitter electrode transmitted to receiver one
        • Could be put in trouser pocket and operated from outside of pocket
        • One benefit is that it can prevent other people from interacting with sensor
      • Configuration #3: Type-B'
        • Same sensor structure as Type-B
        • Differs from Type-B by placement of transmitter and receiver electrodes (reversed)
    • Interaction areas on clothing
      • Several variations of GesturePad attached to different clothes parts
        • Several parts could be assigned as input areas and used in combination
        • For user's convenience:
          • Frequently used commands assigned to easily accessible areas (e.g., lapels, sleeves)
          • Rarely used commands assigned to elbows or knees
        • Traditional devices (e.g., mice)
          • Requires changing operation mode or selecting menu item before actually performing operation
          • When interacting with wearable computers, this step is troublesome
    • Clothes as communication surfaces
      • Another feature of GesturePad is ability to communicate with other digital devices
      • Can typically be applied to control data communication between digital devices nearby

Comments

0 comments:

Post a Comment