Eye position distribution depending on head orientation in watching Ultra High Definition Television Yu Fang (fangyork@riec.tohoku.ac.jp) 1 , Masaki Emoto 2 , Ryoichi Nakashima 3 ,4 , Kazumichi Matsumiya 3 , Ichiro Kuriki 3 , Satoshi Shioiri 3 ,4 1 Graduate School of Information Sciences, Tohoku University 2 NHK Science & Technology Research Laboratories 3 Research Institute of Electrical Communication,Tohoku University 4 Japan Science and Technology Agency, CREST D3-014 BACKGROUND MATERIALS & METHODS RESULTS: Vertical PURPOSE RESULTS: Horizontal CONCLUSION 110cm 189cm • Ultra High-Definition Television • View in a chair • Distance:55cm & 110cm • Stimulus: 15 minutes video clip • Task: to detect a scene change and to detect blur in the movie. EMR-9 FASTRAK • 20 participants Eye position distribution depending on head orientation Visual search (360° display: viewing distance 60 cm) Horizontal head orientation biases eye position toward the same direction with head (Fang et al., VSS2012). Viewing Distance:about 60 cm (Fang et al., VSS2012) Viewing Distance:about 136 cm (Nakashima et al., APCV2012) Static natural scene viewing (56 deg x 42 deg: viewing distance 136 cm) Horizontal head orientation biases eye position toward the same direction with head; Vertical head orientation does not influence eye position distribution. (Nakashima et al., APCV2012) To investigate the eye-head coordination while viewing movies in a high-resolution display (189 x 110 cm). To investigate whether the viewing distances/size can influence the eye-head coordination. Typical eye, head and gaze movements. • Eye position (relative to the head) • Head orientation (relative to the chest) • Gaze = Eye + Head (55㎝) (110㎝) ~120 deg ~80 deg Head orientation biases eye position toward the same direction with head. Head orientation effect on eye distribution (slope) was not different significantly between the two viewing distances (paired t-test, t(19)=1.18, p=0.25) ; Head moved more widely (standard deviation of head distribution) with a 55 cm viewing distance than with a 110 cm distance (paired t-test, t(19)=5.76, p<.001). (55㎝) (110㎝) -6 -3 0 3 6 Eye position Frequency -15 15 Eye position Frequency -6 -3 0 3 6 -15 15 Horizontal: Head orientation effect on eye distribution (slope) in the visual search experiment with 360 degree display is significantly larger than that in the others. Vertical: Slope in the present study with movies is significantly larger than in the other experiments with static stimuli. References • Fang, Y., Nakashima, R., Matsumiya, K., Tokunaga, R., Kuriki, I., Shioiri, S. (2012). Eye position distribution depends on head orientation. VSS, Naples, USA. • Nakashima, R., Fang, Y., Matsumiya, K., Tokunaga, R., Kuriki, I., Shioiri, S. (2012). Eye position distribution depends on head orientation in natural scene viewing. APCV, Incheon, Korea. We found similar eye-head coordination effect with high resolution movies to that with static images in the horizontal direction. The eye-head coordination process is suggested to function even in the vertical direction in the present experiment, inconsistently with in previous studies. The head orientation effect on eye distribution was not different significantly between the two viewing distances. There are possible reasons why the present result differed from the previous ones: • Movie vs. stationary scene • The viewing angle in vertical. • Ultra High-Definition TV with high resolution. Head orientation biases eye position toward the same direction with head as for the horizontal direction, inconsistently with static stimuli. Head orientation effect on eye distribution (slope) was not different significantly between the two viewing distances (paired t-test, t(19)=1.01, p = 0.33) ; Head moved more widely (standard deviation of head distribution) with a 55 cm viewing distance than with a 110 cm distance (paired t-test, t(19)=4.35, p<.001). RESULTS: Head effect Eye position Frequency -6 -3 0 3 6 -15 15 Eye position Frequency -6 -3 0 3 6 -15 15 Head orientation Head orientation Head orientation Head orientation