Yingzi Lin

A driver fatigue recognition model based on information fusion and dynamic Bayesian network

We propose a driver fatigue recognition model based on the dynamic Bayesian network, information fusion and multiple contextual and physiological features. We include features such as the contact physiological features (e.g., ECG and EEG), and apply the first-order Hidden Markov Model to compute the dynamics of the Bayesian network at different time slices. The experimental validation shows the effectiveness of the proposed system; also it indicates that the contact physiological features (especially ECG and EEG) are significant factors for inferring the fatigue state of a driver.

Optic flow and geometric field of view in a driving simulator display

This article reports on a study that used a driving simulator to display conditions of high and low optic flow. The study included 30 drivers who were requested to produce vehicle speeds of either 30 or 60 mph with the geometric field of view at 25, 55, or 85 visual degrees. Drivers overestimated the production of 30 by 20 mph. It was also found that the production of speed was highly dependent on the geometric field of view. These results suggest that optic flow presented in a driving simulator display does not correspond with optic flow found in real-world driving.

Objective and subjective measures of visual aesthetics of website interface design: the two sides of the coin

The main purpose of this study is to compare objective layout-based measures of visual aesthetics with subjective questionnaire-based measures. Correlation analysis was used to carry out the comparison. Values for the tested objective measures were calculated for forty-two web pages already used in a previous study, for which subjective questionnaire scores (classical/expressive and VisA WI) were already available. Results showed significant correlations between many of the tested objective screen layout-based measures and subjective questionnaire-based measures related to order and layout of the screen. These findings suggest that the objective layout-based measures tested in this study can be used for overall assessments of visual aesthetics of websites and particularly for assessing aesthetic aspects related to the classical and the simplicity dimensions of website aesthetics.

Investigating effects of screen layout elements on interface and screen design aesthetics

A recent study suggested the use of the screen layout elements of balance, unity, and sequence as a part of a computational model of interface aesthetics. It is argued that these three elements are the most contributed terms in the model. In the current study, a controlled experiment was designed and conducted to systematically investigate effects of these three elements (balance, unity, and sequence) on the perceived interface aesthetics. Results showed that the three elements have significant effects on the perceived interface aesthetics. Significant interactions were also found among the three elements. A regression model relating the perceived visual aesthetics to the three elements was constructed. When validating the model using standard questionnaire scores of real web pages, high correlations were found between the values computed by the model and scores of questionnaire items related to visual layout of the web pages, indicating that layout-based measures are good at assessing the classical dimension of website aesthetics.

Multimodality inferring of human cognitive states based on integration of neuro-fuzzy network and information fusion techniques

To achieve an effective and safe operation on the machine system where the human interacts with the machine mutually, there is a need for the machine to understand the human state, especially cognitive state, when the humans operation task demands an intensive cognitive activity. Due to a well-known fact with the human being, a highly uncertain cognitive state and behavior as well as expressions or cues, the recent trend to infer the human state is to consider multimodality features of the human operator. In this paper, we present a method for multimodality inferring of human cognitive states by integrating neuro-fuzzy network and information fusion techniques. To demonstrate the effectiveness of this method, we take the driver fatigue detection as an example. The proposed method has, in particular, the following new features. First, human expressions are classified into four categories: (i) casual or contextual feature, (ii) contact feature, (iii) contactless feature, and (iv) performance feature. Second, the fuzzy neural network technique, in particular Takagi-Sugeno-Kang (TSK) model, is employed to cope with uncertain behaviors. Third, the sensor fusion technique, in particular ordered weighted aggregation (OWA), is integrated with the TSK model in such a way that cues are taken as inputs to the TSK model, and then the outputs of the TSK are fused by the OWA which gives outputs corresponding to particular cognitive states under interest (e.g., fatigue). We call this method TSK-OWA. Validation of the TSK-OWA, performed in the Northeastern University vehicle drive simulator, has shown that the proposed method is promising to be a general tool for human cognitive state inferring and a special tool for the driver fatigue detection.

The Effect of Driving Environments on Simulator Sickness

In order to be an effective tool for driver evaluation and education, driving simulators need to be better designed to reduce simulator sickness. This study investigated driving in four environments (country, suburban, city, and curves) using a simulator. When driving on straight roads (city and suburban environments) subjects reported less simulator sickness then driving in the city environment (which included left and right turns) and on curves. A mini-SSQ was used to measure the accumulation of simulator sickness over trials. From trial 1 to trial 5, reported simulator sickness increased linearly. From trial 5 through 8, the rate of increase in simulator sickness decreased. We suggest that the rapid and distorted optic flow experienced while executing turns and driving on curves in driving simulators makes a substantial contribution to simulator sickness.

Finger-based multitouch interface for performing 3D CAD operations

The area of multitouch interaction research is at its infancy. The commercial sector has seen an exponential growth in this area with ubiquitous products like Apple i-Phone, i-Pad, and Microsoft surface table. In spite of their popularity, developers are still finding it difficult to extend this novel interface to engineering applications such as computer aided design (CAD), due to insufficient understanding of the factors that affect the multitouch interface interaction when applied to CAD operations. The objective of this research is to (1) outline the key elements of the multitouch interface for CAD, (2) identify the factors affecting the performance of a multitouch enabled CAD modeling environment, and (3) lay a foundation for future research and highlight the directions for extending the multitouch interface for CAD and other engineering applications. To demonstrate specific results we have conducted mouse emulation experiments. We compared the performance of two finger touch-based interaction techniques (drag state finger touch and track state finger touch) and a standard mouse device for 3D CAD modeling operations. The results indicated that both the task completion time and error rates are statistically the same for both the finger touch-based techniques. However, the error concentration observed from the experiments revealed that for the edge selection tasks, the track state technique is more suited than the drag state technique. Both the finger touch-based techniques suffered from precise dimension control while executing the tasks. The inclusion of a grid on the design space for modeling purpose reduced user errors. The mouse device outperformed both the finger touch-based techniques and yielded statistically better results in terms of task completion time and error rates.

A multiple autocorrelation analysis method for motor imagery EEG feature extraction

A novel multiple autocorrelation method for single trial EEG feature extraction was proposed. The time courses of ERD/ERS during motor imagery were investigated by calculating multiple autocorrelation before power spectrum analysis. Then the averaged power spectrums on specific frequency bands were sent to a K-nearest classifier to validate the separability between different classes. Compared with the result of power spectrum, the multiple autocorrelation performed better in attenuating noise and enhancing the separability between different classes with a small quantity of electrodes (C3 and C4). The maximum 90.0% accuracy tested on dataset of BCI-competition 2003 for motor imagery is achieved.

Effects of Unity of Form and Symmetry on Visual Aesthetics of Website Interface Design

Unity of form represents the extent to which visual objects on the screen are related in size. High levels of unity of form can be achieved by using objects with similar sizes on the screen and/or by reducing number of objects on the screen. Findings of earlier observational studies suggested that effect of unity of form on perceived visual aesthetics of website design was more evident in case of highly symmetrical webpages. The purpose of this study is to verify these findings. An experiment was conducted to systematically study effects of number of objects and number of different sizes of objects on perceived visual aesthetics of website design at high and low levels of symmetry. Perceived aesthetics was assessed using the classical/expressive dimensions. Results showed that both factors have significant effects on perceived visual aesthetics, only at high levels of symmetry and only on the classical dimension.

Gaze tracking system for teleoperation

Conventional control interfaces such as joystick, keyboard, and mouse are widely used in teleoperation. However, they are not operated by some users such as who have disabled hands. This paper proposed an eye gaze tracking system for teleoperation. The eye gaze tracking system is based on non-intrusive, visible light, iris center corneal reflection (ICCR) and 2-D mapping-based gaze estimation. In addition, this paper also presented a simple and effective 2-D mapping-based calibration method in order to users with head movement. The calibration verification of five users shows the average errors of our proposed gaze tracker on the x- and y directions are 13.3 and 16.4 pixels, respectively. Through the teleoperation test of robotic car within fixed experimental area, the gaze tracking system achieved a good control result.