Page 87 - untitled

࿋̾๕
Liao, Hong-Yuan Mark
• Professor, National Chiao Tung University
(
(਷ͭʹஷɽኪΥ໌઺બ, 2003-present)
• Assistant Research Fellow, Academia Sinica
(1991-1995)
• Associate Research Fellow, Academia Sinica
(1995-1998)
• Young Investigator§s Award, Academia Sinica
(1998) (ʕ̯޼Ӻ৫ϋჀኪ٫ᆤ)
• Distinguished Research Award, National Science
Council (2003) €਷߅ึ௫̈޼Ӻᆤ
• National Invention Award( 2004) (਷࢕೯׼ᆤ)
Associate Editor:
• IEEE Transactions on Multimedia (1998-2001)
• Journal of Visual Communication and Image Repre-
sentation (Academic Press) (2000~present)
• Acta Automatica Sinica (2000~2005)
• Managing Editor, Journal of Information Science
and Engineering (2001~present)
• EURASIP Journal on Applied Signal Processing
(2004~present)
• Journal of Computers (2006.3~2008.9)
76
ਿ ͉ ༟ ࣘ Research Description
ਿ
Research Description
ࣘ
͉
༟
ᔖcc၈j ޼Ӻࡰࡒʕ޼৫ࠇၑʕː˴΂Research Fellow My current research interests include: the In the area of visual salience-guided 3D mesh
(1998-) development of video-based surveillance systems decomposition/retrieval, we have developed a visual
and the design of visual salience-guided 3D mesh salience-guided mesh decomposition algorithm.
௰৷ኪዝj ߕ਷Г̏ɽኪཥዚ௹ɻPh.D Electrical retrieval systems. The key concept of this algorithm is to base the de-
Engineering, Northwestern University (1990) composition process on the theory of part salience.
The research on video-based surveillance sys-
ཥcc༑j+886-2-2788- 3799 ext.1550, 1519
tems is concerned with the development of efficient The theory is originated from cognitive psy-
ෂccॆj+886-2-2789-3309 human motion analysis algorithms and the design of chology and it asserts the salience of a part is de- Research Fellows
abnormal behavior detection algorithms. We have termined by the boundary strength, the degree of
ཥɿڦᇌjliao@iis.sinica.edu.tw
developed an efficient comparison algorithm which protrusion, and the relative size of a part. We have
ၣccࠫjhttp://www.iis.sinica.edu.tw/pages/liao is able to compare two arbitrary human motion tra- successfully converted the above three concepts into
jectories. The above mentioned trajectories may be computational processes and used them to guide a
calculated and derived from some complicated hu- top-down 3D mesh decomposition process. Having
man articulated motions. With the above underlying a decomposed 3D mesh labeled with part salience,
Research Fellows
techniques, one can perform on-line event detection we are able to conduct a quick 3D mesh retrieval
(spatial domain) or off-line event search from a through the network.
޼Ӻᔊʧ
޼Ӻᔊʧ
large compressed video database.
Ңͦۃٙ޼Ӻࠠːܼ̍j೯࢝˸ൖৃމਿᓾٙ
္છӻ୕ʿணࠇ˸Ⴉٝːଣኪɪൖᙂࠠࠅ׌މਿᓾ
ٙɧၪي΁ฤరӻ୕f
ί˸ൖৃމਿᓾ္ٙછӻ୕˙ࠦdҢࡁٙ޼Ӻ
Selected Publications
Selected Publications
ࠠː׳ί೯࢝ϞࣖଟٙɛᗳБމʱؓစၑجۆʿண
ࠇʔ͍੬Бމٙਈ಻စၑجfͦۃҢࡁʊ೯࢝ɓࡈ 1. C.-S. Lu, , S.-K. Huang, C.-J. Sze, H.-Y. Mark Liao "Cocktail water- 7. C. -C Shih, H.-Y. Mark Liao, and C.-S Lu "A New Iterated Two-band
marking for digital image protection," IEEE Trans. on Multimedia, Diffusion Equation :Theory and Its Application," IEEE Trans. on Im-
Ϟࣖٙစၑجd̴̙͜Ըˢ࿁΂จՇૢɛᗳ༶ਗБ
volume 2, number 4, pages 209-224, December 2000. age Processing, volume 12, number 4, pages 466-476, January 2003.
މהପ͛ٙᔊʷࠐ༦fவԬᔊʷٙࠐ༦ԫྼɪ݊͟ 2. C.-J. Sze, H.-Y. Mark Liao, and K.-C. Fan "A new image fl ux conduc- 8. Y.-M. Liang, H.-R. Tyan, H.-Y. Mark Liao, and S.-W. Chen "Video
tion method and its application to selective image smoothing," IEEE Stabilization for a Camcoder Mounted on a Moving Vehicle,"
ɛᗳልᕏٙ৷ܓືٮ༶ਗהપစ̈ԸfϞəɪࠑٙ
Trans. on Image Processing, volume 10, number 2, pages 296-306, IEEE Trans. on Vehicular Technology, volume 53, number 6, pag-
ਿ͉ࠐ༦ˢ࿁စၑجdшɛу̙ੂБуࣛԫ΁ਈ಻ February 2001. es 1636-1648, November 2004.
3. C.-S Lu, and H.-Y. Mark Liao "Multipurpose Watermarking for Image 9. H.-Y. Lin, H.-Y. Mark Liao, C.-S. Lu, and J.-C. Lin "Fragile Water-
א٫ίᏀᐵਹ০࿁ɽඎԫ΁ٙᏀᐵൖৃ˪ݬආБҞ
Authentication and Protection," IEEE Trans. on Image Processing, marking for Authenticating 3-D Polygonal Meshes," IEEE Trans. on
஺ฤరf volume 10, number 10, pages 1579-1592, October 2001. Multimedila, volume 7, number 6, pages 997-1006, December 2005.
4. L.-F.Chen, H.-Y. Mark Liao and J.-C. Lin "Wavelet-based Optical 10. C.-W Su, H.-Y. Mark Liao, H.-R. Tyan, K.-C. Fan, and L.-H Chen "A
ί˸ൖᙂࠠࠅ׌މਿᓾהኬˏٙɧၪي΁ᓘ՟ Flow Estimation," IEEE Trans. on Circuits and Systems for Video Motion-Tolerant Dissolve Detection Algorithm," IEEE Trans. on Mul-
Technology, volume 12, number 1, pages 1-12, January 2002. timedila, volume 7, number 6, pages 1106-1113, December 2005.
˙ࠦdҢࡁίཀ̘வݬࣛගʊ೯࢝̈ɓࡈ˸ൖᙂࠠ 5. C.-S Lu, H.-Y. Mark Liao, and Martin Kutter "Denoising and Copy 11. Y.H. Ho,C.W.Lin, J.F. Chen and H.Y. Mark Liao "Fast Coarse-to-Fine
Attacks and Resilient Watermarking by Exploiting Knowledge at Video Retrieval Using shot-level Spatio-temporal Statistics," IEEE
ࠅ׌ˏኬي΁ʱ௲ٙစၑجfவࡈစၑجٙਿᓾ݊
Detector," IEEE Trans. on Image Processing, volume 11, number 3, Trans. on Circuits and Systems for Video Technology, volume 16,
л͜ϓʱي΁ٙࠠࠅ׌jܼ̍ϓʱي΁ၾ˴᜗ʘග pages 280-292, March 2002. number 5, pages 642-648, May 2006.
ٙᗙޢ੶ܓdவي΁ٙ߉̈೻ܓd˸ʿϤϓʱي΁ 6. C.-S Lu and H.-Y. Mark Liao "Structural Digital Signature for Image 12. H.-Y. Sean Lin, H.-Y. Mark Liao, and J.-C. Lin "Visual Salience Guid-
Authentication: An Incidental resistant Scheme," IEEE Trans. on Mul- ed Mesh Decomposition,"To Appear in IEEE Trans. on Multimedila.
ٙɽʃˉʂഃdԸኬˏي΁ʱ௲fҢࡁϓ̌ήਗ਼ɪ timedila, volume 5, number 2, pages 161-173, June 2003.
ࠑ฿ׂᔷ౬ϓࠇၑجۆԸੂБɓࡈ͟ɪϾɨٙϞࣖ
ʱ௲fϞəɪࠑٙϞࣖʱ௲dшɛу̙л͜ˇඎࠠ
ࠅٙϓʱي΁ԸੂБҞ஺ɧၪي΁ᓘ՟f
77