Galaxy Nexus Save Overall Analyze Settings Google, Swipe Speed Facebook LINE Speed (inch/s) Time PC! Android! OS OS

Transcription

1 情報処理学会インタラクション 0 IPSJ Interaction 0 A0-5 0//7 TouchContext:,a) GPS Analayzer TouchContext: Human Context Recognition based on Touch Operation Analysis Yuko Hirabe,a) Yutaka Arakawa Keichi Yasumoto Abstract: This paper proposes to use touch operation for recognizing a human context as one of available sensors in a smartphone. We have developped a system for logging and analyzing all the touch operation on a smartphone. And we report some evaluation results that show a possibility to apply a touch context for recognizing a human context.. [][] GPS Global Positioning System GPS WiFi Nara Institute of Science and Technology 896-5, Takayama, Ikoma, Nara 60 09, Japan a) hirabe.yuko.ho@is.naist.jp [] [5] OS Operating System UI User Interface Android OS Operating System Sandbox 0 Information Processing Society of Japan 58

2 Galaxy Nexus Save Overall Analyze Settings Google, Swipe Speed Facebook LINE Speed (inch/s) Time PC! Android! OS OS Android OS. Web ClickTale * Web ClickTale Touch ClickTale * TouchAnalyzer TouchLogger[6] Touchalytics[7] USB!. TouchAnalyzer PC USB TouchAnalyzer Android SDK platform-tools Android Debug Bridge adb python TouchAnalyzer AndroidOS AndroidOS Information Processing Society of Japan 59

3 . Android Linux /dev/input TouchAnalyzer /dev/input adb foreground () /dev/input event0 event* * event. event -. 6 Android OS Galaxy Nexus.. Galaxy S III.0. Ideapad Tablet.. Xperia Arc HD.0. 5 Nexus.. 6 Galaxy Note II.. 7 GalaxyS II.0. Android.0. GalaxyS II. /dev/input TouchAnalyzer ==== ==== ==== 000 (a) (g) Semicolon + space Space Space : bd Elapsed time from wake-up Time format or Flag of processing Type 0000 : unprocessed 000 : during processing 7 Value 0000 : Delimiter 00f : (a) Tracking number of recognized finger 009 : (b) Sequence number 005 : (c) x-coordinate 006 : (d) y-coordinate 000 : (e) Radius of ground contact area 00a : (f) Pressure of ground contact area 009 : (g)touch exit 0 Information Processing Society of Japan 60

4 00f (a) 009 (b) (g) ffffffff(g) (b) 005 (c) x 006 (d) y 000 (e) GalaxyS II WVGA (Wide-VGA) 800px 80px [rad px ] 00a (f) [hpa] f (a) b (b) (c) x ca (d) y (e) ==== ==== ==== f (a) b (b) c (c) x b (d) y d (e) a (f) ==== ==== ==== f (a) ffffffff (g) ==== ==== ==== f (a) ffffffff (g) ==== ==== ==== /dev/input/event 7 (a) (g) (a) (g) (b) (c) (d) (g) (a) (e) (f) (a) (g).. (a) (a) (a) (b) (f) (a) Algorithm Algorithm if (a) then (b) (f) else {(a) } for (a) && 0 do (b) (f) end for end if (b) (f) 0 TouchAnalyzer 8.. Algorithm TouchAnalyzer 9.. P P L P P L L L TouchAnalyzer 0..5 Algorithm P P P L P P P L L L 0 Information Processing Society of Japan 6

5 B(x 0, y n ) (x k, y k ) Time T k Time T n A(x 0, y 0 ) Time T 0 8 x ab = (xn x 0 ) + (y n y 0 ) T n T 0 () A B x TouchAnalyzer n (xk+ x k ) + (y k+ y k ) 9 x ab = n k=0 T k+ T k n () P P u P P v P P w L L TouchAnalyzer 0 Starting point P Starting point P w L End point P u End point P K L K End point P End point P v Starting point P Starting point P 0.. TouchAnalyzer PC USB y TouchAnalyzer foreground. 0 Information Processing Society of Japan 6

6 The counted number of each touch area Subject Counted Numbers 0 A number of divided areas - width A numbe of divided areas - height The counted number of each touch area Subject! Counted Numbers! A number of divided areas - width! A numbe of divided areas - height 5 Galaxy SII Facebook. Yahoo QWERTY 0 Information Processing Society of Japan 6

7 ( 7) ( 8) ( 9) ( 0). Facebook Facebook MSRA 5. 0 Information Processing Society of Japan 6

8 9 0 UI [] GPS Log. [] NIKON IMAGE SPACE. [] Kawaguchi, N., Ogawa, N., Iwasaki, Y., Kaji, K., Terada, T., Murao, K., Inoue, S., Kawahara, Y., Sumi, Y. and Nishio, N.: HASC Challenge: gathering large scale human activity corpus for the real-world activity understandings, Proceedings of the nd Augmented Human International Conference, ACM, p. 7 (0). [] Kawaguchi, N., Ogawa, N., Iwasaki, Y., Kaji, K., Terada, T., Murao, K., Inoue, S., Kawahara, Y., Sumi, Y. and Nishio, N.: HASC Challenge: gathering large scale human activity corpus for the real-world activity understandings, Proceedings of the nd Augmented Human International Conference, ACM, p. 7 (0). [5] Fukumoto, K., Terada, T. and Tsukamoto, M.: A smile/laughter recognition mechanism for smile-based life logging, Proceedings of the th Augmented Human International Conference, ACM, pp. 0 (0). [6] Cai, L. and Chen, H.: TouchLogger: inferring keystrokes on touch screen from smartphone motion, Proceedings of the 6th USENIX conference on Hot topics in security, USENIX Association, pp. 9 9 (0). [7] Frank, M., Biedert, R., Ma, E., Martinovic, I. and Song, D.: Touchalytics: On the applicability of touchscreen input as a behavioral biometric for continuous authentication (0). 0 Information Processing Society of Japan 65