Browse Prior Art Database

Method and System for Sensing Group State based on a Combination of Wearable Sensor Data Disclosure Number: IPCOM000243492D
Publication Date: 2015-Sep-24
Document File: 3 page(s) / 65K

Publishing Venue

The Prior Art Database


A method and system is disclosed for sensing group state based on a combination of wearable sensor data.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 3

Method and System for Sensing Group State based on a Combination of Wearable Sensor Data

Continuous engagement during unicast group meetings are easily broken by distractions, changes in individual state such as stress, cognitive load, and environment or unintentional actions of a presenter. There is a need for methodologies that help capture the state of a crowd, detects suboptimal conditions and reacts accordingly. Existing approaches capture group state for unicast meetings by employing real -time facial image capture and analysis of images for emotion . These approaches are expensive, ineffective for moving subjects and fail to capture physiological states related to inattention for individuals with flat facial expression response .

Disclosed is a method and system for sensing group state based on a combination of

wearable sensor data. The method and system utilizes selective sampling and combination of wearable sensor data to understand group state and adapt group -based interfaces. The method and system utilizes inexpensive wearable sensors for group state sensing and group interface adaptation. Rule based priority weights are applied to sensor data for estimating group state. The method and system uses algorithms and sensor data to detect a set of human activities relevant to unicast meetings such as typing, mouse use, wave gestures and hand raised gestures. Thereafter, visualization methods are utilized for understanding group states and communicating meaning .

The method and system is illustrated in the figure .


As illustrated in the figure, an attention level module computes a composite attention score for each measured individual within a group based on known gestural movement and general movements such as fidgeting or gesticulations . Thereafter, movement


Page 02 of 3

data is estimated from sensor accelerometer and gyroscope , and each sensor is calibrated to individual baselines to ensure accuracy.

An autonomic stress and cognitive load module computes a composite cognitive load and stress score for each measured individual based on heart rate variability sensors and galvanic skin response sensors respectively. A weighted data integration module analyzes physiological responses from the attention and stress modules and computes Rule Based Priority Weights (RBPWs) for each score used in determining crowd state metrics. RBPWs are learned from analyzing group data and may also be constructed by explicit rules. For example, higher weights may be assigned for important members and lower weights for individuals with lesser impact. The weight assignment process is an important quality control because different individuals may contribute differently to an overall crowd state. The weight assignment process also ensures crowd state estimates are robust to differences in individual physiological respons...