Overview
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore a groundbreaking gesture recognition system in this 20-minute conference talk from NSDI '14. Discover AllSee, a low-power solution that operates on various devices, including those without batteries. Learn how it extracts gesture information from existing wireless signals, consuming significantly less power than current systems. Examine the prototype's ability to recognize gestures on RFID tags and power-harvesting sensors, as well as its integration with smartphones for through-the-pocket scenarios. Delve into the system's architecture, challenges overcome, and impressive classification accuracies across eight gestures. Gain insights into the potential of this technology for always-on gesture recognition in smartphones, tablets, and Internet of Things devices.
Syllabus
Intro
New Forms of User Interaction
Limitations of Connect
Internet of Things Devices
AlwaysOn
HighLevel Idea
AllC
Hardware Architecture
Wireless Receiver
Challenges
Ambient Backscatter
Power Consumption
Design Challenges
Demonstration
Gesture Classification
Workflow
Classification Accuracy
False Positives
Real World Testing
Conclusion
Taught by
USENIX