The Joint Seminar on Computational Intelligence by IEEE CIS Thailand Chapter provides the opportunities for researchers from member institutes to share their work on computational intelligence in a friendly and constructive atmosphere. Some interesting emerging issues will also be discussed by guest speaker.
School of Information Technology, King Mongkut’s University of Technology Thonburi will host the 2nd Joint Seminar on Computational Intelligence on Thursday 23rd February 2017. There will be two talks by distinguished guest speakers and six presentations by students and researchers from KMITL, KMUTNB, and KMUTT.
- Prof. Dr. Mark Chignell, University of Toronto, Canada — “Gait-Guided Adaptive Interfaces for Dismounted Soldiers”
Registration period: 17-21 February 2017, 03.00 PM
Register at: https://docs.google.com/forms/d/e/1FAIpQLSdvcBy93MFtedq_JalWMgu0VA2Xzbw7kDddRO8ywsMcObk1Yw/viewform
For more information, please contact Junjira Tumchaiporn (email@example.com)
Click here > [PDF]
|08:00 – 09:00
|09:00 – 10:15
||Invited Speaker – Gait-Guided Adaptive Interfaces for Dismounted Soldiers, Prof. Dr. Mark Chignell, University of Toronto, Canada [PDF]
|10:15 – 10:30
|10:30 – 11:00
||Detecting Falls, Activities of Daily Living and Indoor Positioning using Smartwatch and Beacons, by Sitthichai Sukreep, King Mongkut’s University of Technology Thonburi (KMUTT) [PDF]
|11:00 – 11:30
||The Performance Evaluation of a Hybrid Immune Genetic Algorithm Based on Mathematical Functions, by Pongsaran Boonyopakorn, King Mongkut’s University of Technology North Bangkok (KMUTNB) [PDF]
|11:30 – 12:00
||Distillation of Knowledge from the Research Literatures on Alzheimer’s Dementia, by Wutthipong Kongburan, King Mongkut’s University of Technology Thonburi (KMUTT) [PDF]
|12:00 – 13:00
|13:00 – 14:15
||Invited Speaker – Driver Distraction and Cognitive Workload: Measurement Methods for in-vehicle HMI, Dr. Sachi Mizobuchi, Toronto Rehabilitation Institute, Canada [PDF]
|14:15 – 14:30
|14:30 – 15:00
||Fractal Dimension for Classifying 3D Brain MRI Using Improved Triangle Box-Counting Method, by Yothin Kaewaramsri, and Kuntpong Woraratpanya, King Mongkut’s Institute of Technology Ladkrabang (KMITL) [PDF]
|15:00 – 15:30
||A Trust in Social Networks for Recommendation Systems, by Thanaphon Phukseng, King Mongkut’s University of Technology North Bangkok (KMUTNB) [PDF]
|15:30 – 16:00
||Comparative Study of Machine Learning Techniques for Automatic Product Categorisation, by Chanawee Chavaltada, Kitsuchart Pasupa, and David R. Hardoon (KMITL) [PDF]
- Prof. Dr. Mark Chignell, University of Toronto, Canada
Topic: Gait-Guided Adaptive Interfaces for Dismounted Soldiers
Abstract: Adaptive interfaces with the goal of maintaining mental workload were proposed by Hancock and Chignell in the 1980s. Adaptation to mental workload levels is important in complex contexts, since it can be detrimental to task performance. For instance, excessively high workload can lead to controlled flight into terrain accidents in aviation, and in low workload vigilance tasks such as sentry duty or radar operation it can lead to missed targets. While many techniques have been proposed, mental workload has proven to be challenging to measure in practice, and it is often measured using self report methods, principally with the NASA TLX scale. Recent research has shown a relationship between the way people walk and the cognitive load that they are under, raising the possibility that modifications to gait can be a proxy measure of mental workload when a person is ambulating while performing tasks. In the interactive media lab at the University of Toronto we have developed a method of gait analysis using a smartphone app. We have validated the app gait results using Vicon motion capture as a gold standard comparison. We have also shown, in an experimental study, that modifications to gait due to a cognitively loading task are correlated with executive function ability as measured by the Stroop task. I will present these experimental results and propose that modifications to gait are an indicator of mental workload. I will then introduce gait-guided adaptive interfaces as a solution to the problem of measuring mental workload in mobile contexts, and as a means of guiding adaptation of tasks so as improve both the well-being and the performance of the operator.
- Dr. Sachi Mizobuchi, Toronto Rehabilitation Institute, Canada
Topic: Driver Distraction and Cognitive Workload: Measurement Methods for in-vehicle HMI
Abstract: Rapid diffusion of mobile information devices and in-car information systems has increased the prevalence of distracted driving, where the driver’s attention is not fully directed to the road, with a resulting increased risk to driving safety. While visual-manual distraction impairs driving performance when using a handset, previous studies have also shown that hands-free conversation also impairs driving performance. Researchers have concluded that cognitive distraction, in addition to visual-manual distraction, can have a negative impact on driving safety. Cognitive distraction is difficult to characterize because, unlike visual-manual distraction, it is hard to determine exactly when the driver is being distracted from the primary (driving) task.
In this talk, I am going to address various methods to measure cognitive workload, discuss strength and weakness of each of the methods, and introduce some in-vehicle applications.