Pradipta Biswas

Eye Gaze Controlled MFD for Military Aviation

Multi-Function Displays (MFD) are essential part of glass cockpit of modern military and civilian aircrafts. They can display more information in less space than traditional analog displays. Interacting with MFDs is still now limited to a joystick system attached to throttle (called Target Des-ignation System, TDS). This paper explored using gaze con-trolled interface for MFD and proposed two algorithms based on hotspots and adaptable zooming for improving response times in a gaze controlled interface. Three user studies confirmed gaze controlled MFD can significantly reduce response times for big peripheral buttons compared to touchscreen in a head down configuration and com-pared to existing TDS in a head up configuration.

Gaze Controlled Interface For Limited Mobility Environment

This paper investigates design constraints for developing GUI (Graphical User Interfaces) in limited mobility environments. In particular, we have presented a gaze controlled interface for users with severe physical impairment and for pilots in military aviation who are constrained by situational impairment. We have developed and implemented algorithms to investigate visual search pattern and improve pointing and selection tasks in eye gaze controlled GUI. User studies in both environments show that for situational impaired users, there is improved flying performance and significant reduction in pointing and selection times for secondary mission control tasks compared to existing interaction systems.

Inclusive Personlization of User Interfaces

This paper presents a user modelling system that facilitates developing and using interfaces for people with wide range of abilities. It consists of a simulator that helps designers to understand, visualize and measure effect of physical impairment on user interfaces. The modelling system also consists of a web service to customize static features (like font size, colour contrast) of an interface and a pointing facilitation module that predicts users’ intended target. The paper highlights applications of the system in product design and developing a web based geo-tagging application for reporting civic, healthcare issues and disaster warning messages by elderly users themselves.

User inclusive crowd sourcing disaster reporting system: Reporting and forecast

Inspired by open source software development model of utilizing intelligence and memory of mutually sharable human resources at almost negligible financial cost we develop framework to use perception and cognition of crowd. The paper presents a survey of recent such approaches, associated limitations and scopes of impoverishment. Using inclusive design framework, we design web and android applications while taking care of needs of varied user base. We also propose application for low end phones to cover lower economic and geographic strata. These platforms felicitate both reporting of possibilities i.e predictions/alerts and events occurred in past or ongoing with each event associated with a geo-tag and a point on time line along with other fields. We propose design modifications in platforms for easy and low cognitive tagging.

Preliminary Studies on Input Modalities

This chapter reports a set of user studies on a pointing and selection task similar to ISO 9241 pointing task. Users undertook pointing and selection trials using eye gaze and finger trackers for different target size and distances. We conducted trials in desktop and automotive environments, and results from these trials were later used to develop target prediction and multimodal fusion algorithms.

User Studies on Saccadic Intrusion

The previous chapter confirmed that saccadic intrusion may be a viable alternative to detect higher cognitive load. This chapter presents a set of user studies where participants were subjected to tasks demanding variable amount of cognitive load. We measured and compared parameters involving saccadic intrusion in different task conditions.

Intelligent Multimodal Systems

The previous chapter compared gaze controlled and finger tracking based pointing modalities and we found users found it difficult to home on target using the gaze controlled interface and even with finger tracking system in automotive environment. In this chapter, we have proposed an algorithm that can activate a target even before the pointer reaches on top of it. For a gaze controlled interface, the target will be activated as soon as the saccade launches near the target reducing the fixation duration required to activate a target. In the latter half of the chapter we discussed different fusion strategies to combine eye gaze and finger tracking systems together. The previous chapter already introduced multimodal eye gaze tracking system by combining eye gaze tracking with joystick and LeapMotion controller, this chapter takes forward the concept with more sophisticated fusion models.

User Studies on Driving Simulator

In this chapter, we have investigated the use of eye gaze tracking for operating a dashboard in an automotive environment. In our previous studies, the primary task was pointing and selection in a graphical user interface. In the following two studies, we investigated pointing and selection tasks in a dual-task set-up where the primary task was operating a driving simulator. The first study involved two different secondary tasks to compare four different pointing modalities (touchscreen and the fusion strategies discussed in previous chapters). The second study investigated effect of eye gaze tracking on driving performance in two different road conditions. In the following sections, we have described the studies in details.

Preliminary Study on Cognitive Load Detection

Detection of drivers’ cognitive load is a well-investigated subject in automotive technology. Recent advancement of information technology added a plethora of infotainment systems inside a car to help the driver. However, proper and also improper uses of those systems often distract the driver from driving and lead to driving accidents. Besides the infotainment system, the change of mental states of drivers due to road or traffic condition and duration of driving can also increase chances of accident.