18463906. GESTURE RECOGNITION BASED ON LIKELIHOOD OF INTERACTION simplified abstract (Microsoft Technology Licensing, LLC)

GESTURE RECOGNITION BASED ON LIKELIHOOD OF INTERACTION

Organization Name

Microsoft Technology Licensing, LLC

Inventor(s)

Julia Schwarz of Bellevue WA (US)

Bugra Tekin of Zurich (CH)

Sophie Stellmach of Kirkland WA (US)

Erian Vazquez of Redmond WA (US)

Casey Leon Meekhof of Redmond WA (US)

Fabian Gobel of Zurich (CH)

GESTURE RECOGNITION BASED ON LIKELIHOOD OF INTERACTION - A simplified explanation of the abstract

This abstract first appeared for US patent application 18463906 titled 'GESTURE RECOGNITION BASED ON LIKELIHOOD OF INTERACTION

Simplified Explanation

The patent application describes a method for evaluating gesture input using neural networks. Here are the key points:

• The method involves receiving input data for sequential data frames, including hand tracking data for a user's hands.
• A first neural network is trained to recognize features that indicate subsequent gesture interactions.
• The first neural network evaluates the input data for a sequence of data frames and outputs the likelihood of the user performing gesture interactions within a specific window of data frames.
• A second neural network is trained to recognize features that indicate whether the user is currently performing one or more gesture interactions.
• The second neural network adjusts parameters for gesture interaction recognition based on the indicated likelihood from the first neural network.
• The second neural network evaluates the predetermined window of data frames for performed gesture interactions using the adjusted parameters.
• The second neural network then outputs a signal indicating whether the user is performing one or more gesture interactions during the predetermined window.

Potential applications of this technology:

• Gesture-based user interfaces for electronic devices such as smartphones, tablets, or computers.
• Virtual reality and augmented reality applications that rely on gesture input for interaction.
• Gaming systems that use gestures as a control mechanism.

Problems solved by this technology:

• Accurately recognizing and evaluating gesture interactions in real-time.
• Adjusting parameters for gesture recognition based on the likelihood of gesture interactions.
• Providing a reliable and efficient method for detecting and analyzing gesture input.

Benefits of this technology:

• Improved user experience by allowing natural and intuitive gesture-based interactions.
• Increased accuracy and reliability in recognizing and evaluating gesture input.
• Real-time feedback on whether the user is performing gesture interactions.
• Adaptability to different users and environments by adjusting parameters based on the likelihood of gesture interactions.

Original Abstract Submitted

A method for evaluating gesture input comprises receiving input data for sequential data frames, including hand tracking data for hands of a user. A first neural network is trained to recognize features indicative of subsequent gesture interactions and configured to evaluate input data for a sequence of data frames and to output an indication of a likelihood of the user performing gesture interactions during a predetermined window of data frames. A second neural network is trained to recognize features indicative of whether the user is currently performing one or more gesture interactions and configured to adjust parameters for gesture interaction recognition during the predetermined window based on the indicated likelihood. The second neural network evaluates the predetermined window for performed gesture interactions based on the adjusted parameters, and outputs a signal as to whether the user is performing one or more gesture interactions during the predetermined window.