[image 03630] [講演]明日：7/12京都工芸繊維大学のOPEN TECHシンポジウム

[杜 偉薇]

image-ML$B$N3'MM(B
$B5~ET9)7]A!0]Bg3X!!>pJs9)3X!&?M4V2J3X7O!!;k3P>pJs8&5f<<$NEN$G$9!#(B

$B$*@$OC$K$J$j$^$9!#(B

$BK\Bg3X%9!<%Q!<%0%m!<%P%kBg3XAO ＠ .;Y1g;v6H$K$h$j!":#2s$OCf9qKL5~8rDLBg3X$+$i!"%3%s%T%e!<%?;Y1g?GCG$N8&5f$r$5$l$F$*$i$l$k(BYahui Peng$B;a!"Cf9qE7DEBg3X$+$i!"%S%C%/%G!<%?$NIT3NDj ＠ -%b%G%j%s%0!"%^%k%A%b%@%j%F%#%G!<%?$K$h$k5!3#3X=,!"%$%s%F%j%8%'%s%HL5?M%7%9%F%`$K$D$$$F8&5f$r$5$l$F$*$i$l$k(BQinghua Hu$B;a$r$*>7$-$7!"2<5-$N$H$*$j3+:E$$$?$7$^$9!#$<$R!"$4;22C$r59$7$/$*4j$$CW$7$^$9!#(B

$BF|;~!'NaOB85G/(B7$B7n(B12$BF|!J6b!K(B9$B;~(B30$BJ,!A(B16$B;~(B
$B2q>l!'5~ET9)7]A!0]Bg3X(B 13$B9f4[(B4$B3,B?L\E*<<!!(B

$B!!!!!!9V1i(B
$B!!!!!!!!!!-!(BIntelligent Medical Image Diagnosis
$B!!!!!!!!!!!!9V1i<T!'(BYahui Peng$B!JKL5~8rDLBg3X!&65<x!K!!(B9$B;~(B30$BJ,!A(B11$B;~(B30$BJ,(B
         
$B!!!!!!!!!!-"(BMachine Learning for Intelligent Driving
$B!!!!!!!!!!!!9V1i<T!'(BQinghua Hu$B!JE7DEBg3X!&65<x!K!!!!!!(B14$B;~!A(B16$B;~(B

$B"(K\9V1i$O1Q8l$G9T$$!";vA0?=$79~$_$H;22CHq$,ITMW$G$9!#(B

P.S.
$B-!(BProf. Peng$B$,Cf9q$N ＠ 62ZBg3X$*$h$S%"%a%j%+$N%7%+%4Bg3X$N3X0L$r<hF@$7!"(BComputer aided diagnosis$B!"(BReceiver Operatorating Characteristic curve$B$NJ,Ln$K3hLv$7$F$$$^$9!#(Bhttps://www.researchgate.net/profile/Yahui_Peng
Abstrct: In the presentation, I will introduce how pattern recognition and artificial intelligence techniques were used in a variety of applications of medical imaging and image interpretation, from image segmentation, image registration, feature analysis, classification, and assessment of classification accuracy, in the Image and Signal Processing Laboratory at Beijing Jiaotong University.

$B-"(BProf. Hu$B$,Cf9q$N%O%k%S%s9)6HBg3X$N3X0L$r<hF@$7!"5!3#3X=,$d!"%Q%?!<%sG'<1!"%3%s%T%e!<%?%S%8%g%s$*$h$S(BAI$B$NJ,Ln$K3hLv$7$F$$$^$9!#(B
https://scholar.google.com/citations?user=TVSNq_wAAAAJ&hl=ja
http://cs.tju.edu.cn/faculty/huqinghua/
Abstract:$B!!(BEnvironment sensing and understanding are the key challenging tasks in intelligent unmanned systems. Especially in the open conditions, there are too many variations and uncertainty in modeling the open world, such as weather, location and sunshine. It is still an open problem for unmanned systems to understand the outer conditions. In this talk, I will show some open datasets widely used in this domain, survey some popular techniques based on deep learning and finally describe the work in our lab, including object detection and recognition, image semantic segmentation and multi-sensor fusion.
$B0J>e$G$9!#59$7$/$*4j$$CW$7$^$9!#(B