• Proceedings of the KIEE Conference
• /
• 2005.07d
• /
• pp.2839-2842
• /
• 2005

본 논문에서는 새로운 실시간 시선 추적 방식을 제안하고자한다. 기존의 시선추적 방식은 사용자가 머리를 조금만 움직여도 잘못된 결과를 얻을 수가 있었고 각각의 사용자에 대하여 교정 과정을 수행할 필요가 있었다. 따라서 제안된 시선 추적 방법은 적외선 조명과 Generalized Regression Neural Networks(GRNN)를 이용함으로써 교정 과정 없이 머리의 움직임이 큰 경우에도 견실하고 정확한 시선 추적을 가능하도록 하였다. GRNN을 사용함으로써 매핑기능은 원활하게 할 수 있었고, 머리의 움직임은 시선 매핑 기능에 의해 적절하게 시선추적에 반영되어 얼굴의 움직임이 있는 경우에도 시선추적이 가능토록 하였고, 매핑 기능을 일반화함으로써 각각의 교정과정을 생략 할 수 있게 하여 학습에 참여하지 않은 다른 사용자도 시선 추적을 가능케 하였다. 실험결과 얼굴의 움직임이 있는 경우에는 평균 90% 다른 사용자에 대해서는 평균 85%의 시선 추적 결과를 나타내었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.11a
• /
• pp.1455-1456
• /
• 2013

최근 스마트 이동단말에 장착된 영상 센서를 통해 획득한 영상에서 실시간으로 사용자의 눈을 인식하거나 시선을 추적하여 콘텐츠 제어, 행태 분석 등을 수행하는 기술에 대한 필요성이 증가하고 있다. 기존의 영상 분석 기반 시선 추적 기술은 고가의 시선 추적기를 기반으로 하며, 추적한 시선 좌표를 통해 실제 응시객체를 분석하는 사후 작업이 필요하다. 이에, 본 논문에서는 OpenCV를 기반으로 스마트 이동단말의 전면에 장착된 영상 센서에서 사용자의 눈을 인식하고, 시선을 추적한 후, 이를 실시간으로 시선 좌표와 화면의 콘텐츠 내 객체 영역을 매칭함으로써, 응시객체를 분석하는 스마트 이동단말 기반 시선 추적기를 설계하였다. 본 논문의 시선 추적기는 시선 추적을 수행하기 위해, 사용자에게 5개의 화면 교정점을 제공하고 응시하도록 하여 시선 추적 범위를 측정한다. 또한, 스마트 이동단말에 내장된 자이로스코프 센서를 통해 기울기가 변경될 시, 이를 반영하여 시선 교정을 수행하도록 설계하였다.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.2 no.3
• /
• pp.61-69
• /
• 2009

In this paper, to propose a new approach to real-time eye tracking. Existing methods of tracking the user's attention to the little I move my head was not going to get bad results for each of the users needed to perform the calibration process. Infrared eye tracking methods proposed lighting and Generalized Regression Neural Networks (GRNN) By using the calibration process, the movement of the head is large, even without the reliable and accurate eye tracking, mapping function was to enable each of the calibration process by the generalization can be omitted, did not participate in the study eye other users tracking was possible. Experimental results of facial movements that an average 90% of cases, other users on average 85% of the eye tracking results were shown.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.6 no.2
• /
• pp.195-201
• /
• 2005

This paper describes preliminary results we have obtained in developing a computer vision system based on active IR illumination for real time gaze tracking for interactive graphic display. Unlike most of the existing gaze tracking techniques, which often require assuming a static head to work well and require a cumbersome calibration process for each person, our gaze tracker can perform robust and accurate gaze estimation without calibration and under rather significant head movement. This is made possible by a new gaze calibration procedure that identifies the mapping from pupil parameters to screen coordinates using the Generalized Regression Neural Networks (GRNN). With GRNN, the mapping does not have to be an analytical function and head movement is explicitly accounted for by the gaze mapping function. Furthermore, the mapping function can generalize to other individuals not used in the training. The effectiveness of our gaze tracker is demonstrated by preliminary experiments that involve gaze-contingent interactive graphic display.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.8 no.2
• /
• pp.477-483
• /
• 2004

This paper describes preliminary results we have obtained in developing a computer vision system based on active IR illumination for real time gaze tracking for interactive graphic display. Unlike most of the existing gaze tracking techniques, which often require assuming a static head to work well and require a cumbersome calibration process for each person our gaze tracker can perform robust and accurate gaze estimation without calibration and under rather significant head movement. This is made possible by a new gaze calibration procedure that identifies the mapping from pupil parameters to screen coordinates using the Generalized Regression Neural Networks(GRNN). With GRNN, the mapping does not have to be an analytical function and head movement is explicitly accounted for by the gaze mapping function. Futhermore, the mapping function can generalize to other individuals not used in the training. The effectiveness of our gaze tracker is demonstrated by preliminary experiments that involve gaze-contingent interactive graphic display.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.3C
• /
• pp.125-132
• /
• 2005

This paper describes a computer vision system based on active IR illumination for real-time gaze discrimination system. Unlike most of the existing gaze discrimination techniques, which often require assuming a static head to work well and require a cumbersome calibration process for each person, our gaze discrimination system can perform robust and accurate gaze estimation without calibration and under rather significant head movement. This is made possible by a new gaze calibration procedure that identifies the mapping from pupil parameters to screen coordinates using generalized regression neural networks (GRNNs). With GRNNs, the mapping does not have to be an analytical function and head movement is explicitly accounted for by the gaze mapping function. Futhermore, the mapping function can generalize to other individuals not used in the training. To further improve the gaze estimation accuracy, we employ a reclassification scheme that deals with the classes that tend to be misclassified. This leads to a 10% improvement in classification error. The angular gaze accuracy is about 5°horizontally and 8°vertically. The effectiveness of our gaze tracker is demonstrated by experiments that involve gaze-contingent interactive graphic display.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.3 no.1
• /
• pp.38-46
• /
• 2010

This paper describes a computer vision system based on active IR illumination for real-time gaze discrimination system. Unlike most of the existing gaze discrimination techniques, which often require assuming a static head to work well and require a cumbersome calibration process for each person, our gaze discrimination system can perform robust and accurate gaze estimation without calibration and under rather significant head movement. This is made possible by a new gaze calibration procedure that identifies the mapping from pupil parameters to screen coordinates using generalized regression neural networks (GRNNs). With GRNNs, the mapping does not have to be an analytical function and head movement is explicitly accounted for by the gaze mapping function. Furthermore, the mapping function can generalize to other individuals not used in the training. To further improve the gaze estimation accuracy, we employ a reclassification scheme that deals with the classes that tend to be misclassified. This leads to a 10% improvement in classification error. The angular gaze accuracy is about $5^{\circ}$horizontally and $8^{\circ}$vertically. The effectiveness of our gaze tracker is demonstrated by experiments that involve gaze-contingent interactive graphic display.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.9 no.2
• /
• pp.322-329
• /
• 2005

This paper describes a computer vision system based on active IR illumination for real-time gaze discrimination system. Unlike most of the existing gaze discrimination techniques, which often require assuming a static head to work well and require a cumbersome calibration process for each person, our gaze discrimination system can perform robust and accurate gaze estimation without calibration and under rather significant head movement. This is made possible by a new gaze calibration procedure that identifies the mapping from pupil parameters to screen coordinates using generalized regression neural networks (GRNNS). With GRNNS, the mapping does not have to be an analytical function and head movement is explicitly accounted for by the gaze mapping function. furthermore, the mapping function can generalize to other individuals not used in the training. To further improve the gaze estimation accuracy, we employ a reclassification scheme that deals with the classes that tend to be misclassified. This leads to a 10$\%$ improvement in classification error. The angular gaze accuracy is about $5^{circ}$horizontally and $8^{circ}$vertically. The effectiveness of our gaze tracker is demonstrated by experiments that involve gaze-contingent interactive graphic display.

• Journal of Digital Contents Society
• /
• v.17 no.5
• /
• pp.367-373
• /
• 2016

In this paper, we propose a novel method to improve the accuracy of eye trackers and how to analyze them. The proposed method extracts a user profile information created by extracting gaze coordinates and color information based on the exact pupil information, and then, it maintains a high accuracy in the display. In case that extract the user profile information, the changes of the accuracy for the gaze time also is estimated and the optimum parameter value is extracted. In the experimental results for the accuracy of the gaze detection, the accuracy was low if a user took a short time in a specific point. On the other hand, when taking more than two seconds, the accuracy was measured more than 80 %.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.40 no.6
• /
• pp.141-148
• /
• 2003

This paper proposes a new algorithm that generates the smooth and realistic transition views from one viewpoint to another on the multi-panorama based navigation system. The proposed algorithm is composed of two steps. One is prewarping that aligns the viewing directions of two panoramic images, and the other is the bidirectional disparity morphing(BDM) that generates the intermediate scene from the aligned panoramic images. For prewarping, we compute the phase correlation between two images in order to obtain the information, such as translation, rotation, and scaling. Then we align the viewing directions of two original images using these information. Afterprewarping, we compute the block based disparity vector(DV) and smooth them using two occluding patterns. As we apply these DVs to the BDM, we can generate the elaborate intermediate scene. We make an experiment on the proposed algorithm with some real panoramic images and obtain good quality intermediate scenes.