Interest Operator Scale Space Wavelet SIFT Interest Operator 2 corner detector 1) SUSAN 3) Harris Shi Tomasi Good feat
|
|
|
- ほのか ねごろ
- 7 years ago
- Views:
Transcription
1 Interest Operator Scale Space Wavelet 1 DP 2-1 Interest Operator Scale space Wavelet SIFT DP c /(21)
2 Interest Operator Scale Space Wavelet SIFT Interest Operator 2 corner detector 1) SUSAN 3) Harris Shi Tomasi Good features to 2) track SUSAN Smallest Univalue Segment Assimilating Nucleus S p 0 p g t I(p) p SUSAN R(p 0 ) R(p 0 ) = max(0, g n(p 0 )), n(p 0 ) = c(p 0, p), c(p 0, p) = p S 1 if I(p 0 ) I(p) t 0 if I(p 0 ) I(p) > t Harris J. Shi C. Tomasi sum of squared difference, SSD p W v SSD (2 1) c /(21)
3 S (p) = (I(q) I(q + v)) 2 q W (2 2) 1 I(q) x, y I x (q), I y (q) I(q + v) I(q) + [I x (q) I y (q)] v 2 S (p) = v T W Ix 2 W I x I y W I x I y W Iy 2 v = vt H(p) v (2 3) (2 4) H(p) p Harris Shi Tomasi λ 1, λ 2 Harris M = λ 1 λ 2 κ(λ 1 + λ 2 ) 2 = det(h) κ trace 2 (H) (2 5) κ Shi Tomasi M = min(λ 1, λ 2 ) (2 6) Scale Space cm m nm km A. Witkin 7) scale space 2 1 Gaussian kernel I(x, y) t > 0 L(x, y, t) L : R 2 R + R 6) L t = 0 c /(21)
4 t 2 1 L(x, y, 0) = I(x, y). t > 0 t g t (x, y) = 1 ( 2πt exp x2 + y 2 ) 2t (2 7) (2 8) L(x, y, t) L(x, y, t) = (g t I)(x, y) = g t (u, v)i(x u, y v)dudv. (2 9) t t ( ) L(x, y, t) = x x g t I (x, y) = x g t(u, v)i(x u, y v)dudv (2 10) Wavelet wavelet 13) D. Gabor 12) g λ,θ,ϕ,σ,γ (x, y) = exp ( x 2 + γ 2 y 2 ) ) cos (2π x 2σ 2 λ + ϕ (2 11) (x, y ) = (x cos θ + y sin θ, x sin θ + y cos θ) (2 12) c /(21)
5 電子情報通信学会 知識の森 2 群 2 編 2 章 図 2 2 Gabor フィルタ gλ,θ,ϕ,σ,γ. 左端を標準に 順に σ, λ, θ を変化させた例 を使って信号の局所的な解析を試みた 図 2 2 しかしこの Gabor フィルタは 周波数領 域における広がりが固定されているところに問題があった 一般に L1 及び L2 ノルムが有限 すなわち Z Z Z Z ψ(x, y) 2 dxdy < ψ(x, y) dxdy <, (2 13) を満たす可積分関数 ψ(x, y) をマザーウェーブレットと呼ぶ このとき 1 W s f (x, y) = Z s Z f (u, v) ψ u x v y dudv, s s (2 14) を関数 f (x, y) の連続ウェーブレット変換と呼ぶ A. Grossman と J. Morlet8) は連続ウェーブ レット変換を研究し 最初に地震学上のデータの解析に応用した 図 2 3 Laplacian of Gaussian(左) と Gabor Filter(右) その後 I. Daubechies9) はある程度滑らかでコンパクトな台をもつ つまり値が 0 でない 領域が有界な 正規直交ウェーブレット基底を構成し S. Mallat10, 11) は Y. Meyer とともに 多重解像度解析 multiresolution analysis と呼ばれる ウェーブレット基底を構成する一般 的方法を提案した ウェーブレットは信号処理 画像解析 通信システム レーダ システ ム制御など 多彩な分野で応用されている 特徴点検出で使われる代表的なウェーブレットとして 上記の Gabor フィルタのほかに 次項に述べる SIFT にも使われる Laplacian of Gaussian LoG や Difference of Gaussian 電子情報通信学会 知識ベース c 電子情報通信学会 /(21)
6 DoG (2 8) Laplacian ( ) 2 x + 2 g 2 y 2 t (x, y) (2 15) LoG f (x, y) blob SIFT D. Lowe SIFT Scale-Invariant Feature Transform 5) SIFT SIFT 6) LoG LoG DoG LoG 2 4 X-Y- 3 DoG 2 4 DoG c /(21)
7 DoG D(p) : D 2 x D x D y H = D x D. (2 16) y D 2 x Harris SIFT trace 2 (H)/ det(h) D(p) L m(p) θ(p) : m(p) = (L(p + x) L(p x)) 2 + (L(p + y) L(p y)) 2, (2 17) ( ) L(p + y) L(p y) θ(p) = tan 1. (2 18) L(p + x) L(p x) 36 80% 2 5 SIFT c /(21)
8 SIFT = 128 c /(21)
9 DP Template Matching T M T N T I M I N I 2 6 T I S D (i, j) (i, j) T i j y x template T input I 2 6 cross-correlation S CC (i, j) = (x,y) T I(i + x, j + y)t(x, y) (2 19) T T {[1, M T ] [1, N T ]} S CC I T 180 T FIR matched filter S CC normalized cross-correlation S NCC (i, j) = S CC (x,y) T T 2 (x, y) (x,y) T I 2 (i + x, j + y) (2 20) S CC S NCC T 1 (2 20) T c /(21)
10 D p (i, j) = I(i + x, j + y) T(x, y) p (x,y) T 1/p (2 21) p = 1 p = 2 p = D max (i, j) = max I(i + x, j + y) T(x, y) (2 22) (x,y) T 2 I(i + x, j + y) T(x, y) 0 1 I(i + x, j + y)t(x, y) 0 21) 2 A, B A B 22) A, B 23, 24) MPEG 25) 3 (2 19) S CC I O(M I N I M T N T ) 2 2 a S CC (i, j) M I N I (i, j) S CC (i, j) c /(21)
11 b coarse-to-fine strategy I, T T 2 n O(M I N I M T N T ) 2 4n 2 n+1 I T c SSDA Sequential Similarity Detection Algorithm: SSDA 26, 27) (2 21) p = 1 D 1 (i, j) I(i + x, j + y) T(x, y) θ T D 1 (i, j) M T N T D 1 θ SSDA D 1 D 1 D 1 26) r ( M T N T ) θ r θ r = λ(r + K r) (2 23) K λ T I λ T I 27) d (2 19) (i, j) (i, j) (i, j + 1) (i, j) (i, j + 1) (i, j) (i, j + 1) 28) (i, j) T I T H T (i, j) I M T N T H I,(i, j) c /(21)
12 S H (i, j) = min ( H T (k), H I,(i, j) (k) ) k (2 24) k 0 S H (i, j) 1 S H (i, j ) min (S H(i, j) B, A B ) + A B A (2 25) A, B (i, j ) (i, j) A B A B A B A B A = B A B (2 25) S H (i, j) S H (i, j ) S H (i, j ) S H (i, j ) S H S CC S H S CC e (2 19) S CC T I T I S CC S CC FFT 4 a 29) S CC (i, j) S CC (i, j) 30) b i j /2 H.264/AVC 1/4 29) c /(21)
13 電子情報通信学会 知識の森 2 群 2 編 2 章 c 弾性マッチング 2 パターンをマッチングする際 一方に微小な歪み ずれ があると整合性が急に悪くな る このため マッチングの際 一方のパターンをゴムのように非線形伸縮させ 最も整合 した時点をマッチング結果とすることがある 要するに画像をゴム膜のように変形させなが らマッチングを図る技術であり これは弾性マッチングと呼ばれる 弾性マッチングの実体 は画素対応関係を表現する線形もしくは非線形写像の最適化であり したがって様々な最適 化手法がこれに利用されている31, 32, 33, 34) 5 特徴点を用いたマッチング 以上のテンプレートマッチングはあくまで画像と画像の照合に基づくものであった これ に対し 両方の画像から複数の特徴点を抽出しておき それら特徴点における特徴量に基づ いてマッチングを求める方法がある すなわち画像 I と T それぞれにおいて 節で述 べた SIFT などによる特徴点を多数抽出し 次に T の特徴点それぞれについて それと最も 類似した特徴量をもつ特徴点を I 上に探索することで I と T の間に点対応を求める その 後 その点対応を用いて 最終的な位置合せ すなわち T に対応する I の領域を求める 図 2 7 は SIFT 特徴点 特徴量を用いた 2 画像 I, T のマッチング結果である ここではテ ンプレート T として I の一部を回転させたものを用いている 回転による誤差で T と I の 間には差異があることに注意されたい 同図右が上述の方法で点対応を求めた結果である ただし その類似度が閾値以下のものは省略している 単純な方法ながら 非常に高い精度 でのマッチング実現されている I T Input and template SIFT keypoints keypoint matching result 図 2 7 特徴点を用いたマッチング 位置合せを行うためには この対応結果を用いて T の移動量や回転量の推定を行うこと になる より一般的に いずれかの画像が射影変換を受けている場合は 射影変換パラメー タの推定することになる 最小二乗法によりパラメータ推定を行うことも考えられる しか し 点対応関係は不安定になることもあり 大きく誤った点対応が 全体的な推定結果に悪 影響を与えることもあり得る このため RANSAC35) のようなロバスト推定法が利用される ことも多い 電子情報通信学会 知識ベース c 電子情報通信学会 /(21)
14 DP (c) Dynamic Programming DP DTW Dynamic Time Warping DP 2 2 8(a) DP 2 pattern Y j pattern Y j j=u i pattern X i pattern X i (a) (b) 2 8 DP 36, 37, DP ) I J O(IJ) 39, 40) DNA 43, 44) DP DP R.Bellman 41) 42) 2 X = x 1, x 2,..., x i,..., x I, Y = y 1, y 2,..., y j,..., y J x i y j X i x i Y j y j j = u i (i = 1,..., I) d i (u i ) = x i y ui (2 26) c /(21)
15 minimize F = I d i (u i ), i=1 with respect to u 1,..., u I, subject to 0 u i u i 1 2, u 1 = 1, u I = J (2 27) (i) u i u i 1 (ii) u i u i 1 j = u i du i /di 2 2 DP (2 27) min F = min u 1,...,u I 0 u i u i 1 2 i=1 = min u 2,...,u I 0 u i u i 1 2 {} I d i (u i ) I d i (u i ) + d 2(u 2 ) + i=3 min u 1 0 u 2 u 1 2 d 1 (u 1 ) (2 28) g 2 (u 2 ) = d 2 (u 2 ) + min F = min u 2,...,u I 0 u i u i 1 2 min u 1 0 u 2 u 1 2 d 1 (u 1 ) I d i (u i ) + g 2 (u 2 ) i=3 (2 29) (2 30) u 1 (2 29) min u 1 min F = min u 3,...,u I 0 u i u i 1 2 {} I d i (u i ) + d 3(u 3 ) + i=4 min u 2 0 u 3 u 2 2 g 2 (u 2 ) (2 31) g 3 (u 3 ) = d 3 (u 3 ) + min F = min u 3,...,u I 0 u i u i 1 2 min u 2 0 u 3 u 2 2 g 2 (u 2 ) I d i (u i ) + g 3 (u 3 ) i=4 (2 32) (2 33) c /(21)
16 u 2 g i (u i ) = d i (u i ) + min u i 1 0 u i u i 1 2 g i 1 (u i 1 ) (2 34) min F = min u I g I (u I ) (2 35) F u I = J min F = g I (J) DP DP u 1,..., u i,..., u I (2 34) DP DP DP i = 1 I j 2 8(b) I J DP (2 34) Input: Output: Initialization: X = x 1, x 2,..., x i,..., x I, Y = y 1, y 2,..., y j,..., y J min F: Minimum elastic matching distance between X and Y. u 1,..., u i,..., u I : Optimal elastic matching between X and Y. 1: g i ( j) = and b i ( j) = nil for i, j DP Recursion: 2: g 1 (1) = d 1 (1) = x 1 y 1 3: for i = 2 to I do 4: for j = 1 to J do 5: d i ( j) = x i y j 6: g i ( j) = d i ( j) + min i 1( j ) max( j 2,1) j j 7: b i ( j) = argmin max( j 2,1) j j g i 1 ( j ) Termination: 8: min F = g I (J) 9: u I = J 10: for i = I 1 downto 1 do u i = b i+1 (u i+1 ) 2 9 DP DP DP DP min F u I,..., u i,..., u 1 DP c /(21)
17 J I O(IJ) 3 DP DP i g i (u i ) DP g i (u i ) (u i i) j for j = i ɛ to i + ɛ do ɛ( J) DP 45) DP 46) DNA FASTA BLAST 43, 44) 4 DNA 44) 2 10 DP 2 44) pattern Z pattern Y pattern X c /(21)
18 ) G 1 = (V 1, E 1 ), G 2 = (V 2, E 2 ), V i E i u, v V 1 u, v G 1 (u, v) E f : V 1 V 2 (2 36) (u, v) E 1 = ( f (u), f (v)) E 2 (u, v) E 1 = ( f (u), f (v)) E 2 (2 37) (2 38) f 1 1 u v f (u) f (v) (2 37)(2 38) 2 11 (2 38) (2 37) f f (2 37) f 2 maximum common subgraph, MCS 2 20) c /(21)
19 MCS J.R. Ullman 15) 2 (2 37) error-correcting error-tolerant H. Bunke MCS 17) W.H. Tsai K.S. Fu 16) M. Fischler R. Elschlager 18) relaxation labeling 19) c /(21)
20 1) S.M. Smith and J.M. Brady, SUSAN - a new approach to low level image processing, Int l J Comput. Vision, vol.23, no.1, pp.45-78, ) J. Shi and C. Tomasi, Good Features to Track, Proc. IEEE Comp. Soc. Conf. Comput. Vision and Patt. Recogn., pp , ) C. Harris and M. Stephens, A combined corner and edge detector, Proc. 4th Alvey Vision Conf., pp , ) J. Matas and O. Chum, M. Urba and T. Pajdla, Robust wide baseline stereo from maximally stable extremal regions, Proc. British Machine Vision Conf., pp , ) D. Lowe, Distinctive image features from scale-invariant keypoints, Int l J Comput. Vision, vol.60, no.2, pp , ) Lindeberg, T., Scale-space theory: A basic tool for analysing structures at different scales, J Appl. Stat., vol.21, no.2, pp , ) A. Witkin, Scale-space filtering, Proc. Int l Joint Conf. on Artif. Intell., pp , ) A. Grossman and J. Morlet, Decomposition of Hardy functions into square integrable wavelets of constant shapes, SIAM J. Math. Anal., vol.15, no.4, pp , ) I. Daubechies, Orthonormal Bases of Compactly Supported Wavelets, Comm. Pure Appl. Math., vol.41, pp , ) S. Mallat, Multiresolution Approximations and Wavelet Orthonormal Bases of L 2 (R), Trans. Amer. Math. Soc., vol.315, pp.69-87, ) S. Mallat, A Theory for Multiresolution Signal Decomposition: The Wavelet Representation, IEEE Trans. Patt. Anal. Mach. Intell., vol.11, no.7, pp , ) D. Gabor, Theory of Communication, J. IEE, vol.93, no.26, pp , ) G. Strang and T. Nguyen, Wavelets and Filter Banks, Cambridge Univ. Press, 1996., I, II, ) D. Conte, P. Foggia, C. Sansone and M. Vento, Thirty Years of Graph Matching in Pattern Recognition, Int l J Patt. Recogn., vol.18, no.3, pp , ) J.R. Ullman, An algorithm for subgraph isomorphism, J. Assoc. Comput. Mach., vol.23, pp.31-42, ) W.H. Tsai and K.S. Fu, Error-correcting isomorphisms of attributed relational graphs for pattern analysis, IEEE Trans. Syst. Man Cybern., vol.9, pp , ) H. Bunke, On a relation between graph edit distance and maximum common subgraph, Patt. Recogn. Lett., vol.18, pp , ) M. Fischler and R. Elschlager, The representation and matching of pictorial structures, IEEE Trans. Comput., vol.22, pp.67-92, ) S. Umeyama, An eigendecomposition approach to weighted graph matching problems, IEEE Trans. Patt. Anal. Mach. Intell., vol.10, pp , ) A.P. Ambler, H.G. Barrow, C.M. Brown, R.M. Burstall and R.J. Popplestone, A versatile computercontrolled assembly system, Proc. Int l Joint Conf. on Artif. Intell., pp , ) J.D. Tubbs, A Note on Binary Template Matching, Patt. Recogn., vol.22, no.4, pp , ),,, vol.60, no.3, pp , ) S.S. Beauchemin and J.L. Barron, The Computation of Optical Flow, ACM Comput. Surv., vol.27, no.3, pp , ),, CVIM, 2006-CVIM-152, ),, BP, c /(21)
21 26) D.I. Barnea and H.F. Silverman, A Class of Algorithms for Fast Digital Image Registration, IEEE Trans. Comput., vol.c-21, no.2, pp , ),,, vol.c-21, no.2, pp , ) V.V. Vinod,, (D), vol.j81-d-ii, no.9, pp , ), : 3D,, vol.90, no.8, pp , ) L.G. Brown, A Survey of Image Registration Techniques, ACM Comput. Surv., vol.24, no.4, pp , 1992.,, bit, pp , ) A.K. Jain, Y. Zhong and M. -P. Dubuisson-Jolly, Deformable Template Models: A Review, Signal Process., vol.71, no.2, pp , ) C.A. Gralbey and K.V. Mardia, A Review of Image-Warping Methods, J Appl. Stat., vol.25, no.2, pp , ) H. Lester and S.R. Arridge, A survey of hierarchical non-linear medical image registration, Patt. Recogn., vol.32, no.1, pp , ) S. Uchida and H. Sakoe, A survey of elastic matching techniques for handwritten character recognition, IEICE Trans. Inf. & Syst., vol.e88-d, no.8, pp , ) M.A. Fischler and R.C. Bolles, Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography, Commun. ACM, vol.24, no.6, pp , ),,, vol.27, no.9, pp , ) H. Sakoe and S. Chiba, A Dynamic Programming Algorithm Optimization for Spoken Word Recognition, IEEE Trans. Acoust. Speech and Signal Proc., vol.assp-26, no.1, pp.43-49, ) H. Sakoe, Two-Level DP-Matching Algorithm A Dynamic Programming Based Pattern Matching Algorithm for Continuous Speech Recognition, IEEE Trans. Acoust. Speech and Signal Proc., vol.assp- 27, no.6, pp , ) and,,, vol.30, no.9, pp , ), DP,, vol.prmu , ) R. Bellman and S. Dreyfus, Applied Dynamic Programming, Princeton Univ. Press, ),,, POD, ) R. Durbin, S. Eddy, A. Krogh and G. Mitchison, Biological Sequence Analysis, Cambridge Univ. Press, ),,, ) C. Raphael, Coarse-to-fine dynamic programming, IEEE Trans. Patt. Anal. Mach. Intell., vol.23, no.2, pp , ), DP, (D), vol.j90-d, no.8, pp , c /(21)
Duplicate Near Duplicate Intact Partial Copy Original Image Near Partial Copy Near Partial Copy with a background (a) (b) 2 1 [6] SIFT SIFT SIF
![Duplicate Near Duplicate Intact Partial Copy Original Image Near Partial Copy Near Partial Copy with a background (a) (b) 2 1 [6] SIFT SIFT SIF](/thumbs/82/86406952.jpg "Duplicate Near Duplicate Intact Partial Copy Original Image Near Partial Copy Near Partial Copy with a background (a) (b) 2 1 [6] SIFT SIFT SIF")
Partial Copy Detection of Line Drawings from a Large-Scale Database Weihan Sun, Koichi Kise Graduate School of Engineering, Osaka Prefecture University E-mail: sunweihan@m.cs.osakafu-u.ac.jp, kise@cs.osakafu-u.ac.jp
Optical Flow t t + δt 1 Motion Field 3 3 1) 2) 3) Lucas-Kanade 4) 1 t (x, y) I(x, y, t)
http://wwwieice-hbkborg/ 2 2 4 2 -- 2 4 2010 9 3 3 4-1 Lucas-Kanade 4-2 Mean Shift 3 4-3 2 c 2013 1/(18) http://wwwieice-hbkborg/ 2 2 4 2 -- 2 -- 4 4--1 2010 9 4--1--1 Optical Flow t t + δt 1 Motion Field
yoo_graduation_thesis.dvi
200 3 A Graduation Thesis of College of Engineering, Chubu University Keypoint Matching of Range Data from Features of Shape and Appearance Yohsuke Murai 1 1 2 2.5D 3 2.1 : : : : : : : : : : : : : : :
IPSJ SIG Technical Report Vol.2010-MPS-77 No /3/5 VR SIFT Virtual View Generation in Hallway of Cybercity Buildings from Video Sequen
VR 1 1 1 1 1 SIFT Virtual View Generation in Hallway of Cybercity Buildings from Video Sequences Sachiyo Yoshida, 1 Masami Takata 1 and Joe Kaduki 1 Appearance of Three-dimensional (3D) building model
(a) (b) (c) Canny (d) 1 ( x α, y α ) 3 (x α, y α ) (a) A 2 + B 2 + C 2 + D 2 + E 2 + F 2 = 1 (3) u ξ α u (A, B, C, D, E, F ) (4) ξ α (x 2 α, 2x α y α,
[II] Optimization Computation for 3-D Understanding of Images [II]: Ellipse Fitting 1. (1) 2. (2) (edge detection) (edge) (zero-crossing) Canny (Canny operator) (3) 1(a) [I] [II] [III] [IV ] E-mail sugaya@iim.ics.tut.ac.jp
光学
Fundamentals of Projector-Camera Systems and Their Calibration Methods Takayuki OKATANI To make the images projected by projector s appear as desired, it is e ective and sometimes an only choice to capture
(MIRU2008) HOG Histograms of Oriented Gradients (HOG)
(MIRU2008) 2008 7 HOG - - E-mail: katsu0920@me.cs.scitec.kobe-u.ac.jp, {takigu,ariki}@kobe-u.ac.jp Histograms of Oriented Gradients (HOG) HOG Shape Contexts HOG 5.5 Histograms of Oriented Gradients D Human
untitled
(Robot Vision) Vision ( (computer) Machine VisionComputer Vision ( ) ( ) ( ) ( ) ( ) 1 DTV 2 DTV D 3 ( ( ( ( ( DTV D 4 () 5 A B C D E F G H I A B C D E F G H I I = A + D + G - C - F - I J = A + B + C -
2.2 6).,.,.,. Yang, 7).,,.,,. 2.3 SIFT SIFT (Scale-Invariant Feature Transform) 8).,. SIFT,,. SIFT, Mean-Shift 9)., SIFT,., SIFT,. 3.,.,,,,,.,,,., 1,
1 1 2,,.,.,,, SIFT.,,. Pitching Motion Analysis Using Image Processing Shinya Kasahara, 1 Issei Fujishiro 1 and Yoshio Ohno 2 At present, analysis of pitching motion from baseball videos is timeconsuming
第25回信号処理シンポジウム 2010年11月24日 26日(奈良) 高精度な画像マッチング手法の検討 A Study of a High-Accuracy Image Matching Method 伊藤 康一 東北大学 高橋 徹 青木 孝文 大学院情報科学研究科 Koichi ITO Toru
第5回信号処理シンポジウム 年月日 日(奈良) 高精度な画像マッチング手法の検討 A Study of a High- Image Matching Method 伊藤 康一 東北大学 高橋 徹 青木 孝文 大学院情報科学研究科 Koichi ITO Toru TAKAHASHI Takafumi AOKI Graduate School of Information Sciences, Tohoku
258 5) GPS 1 GPS 6) GPS DP 7) 8) 10) GPS GPS 2 3 4 5 2. 2.1 3 1) GPS Global Positioning System
Vol. 52 No. 1 257 268 (Jan. 2011) 1 2, 1 1 measurement. In this paper, a dynamic road map making system is proposed. The proposition system uses probe-cars which has an in-vehicle camera and a GPS receiver.
IPSJ SIG Technical Report Vol.2010-CVIM-170 No /1/ Visual Recognition of Wire Harnesses for Automated Wiring Masaki Yoneda, 1 Ta
1 1 1 1 2 1. Visual Recognition of Wire Harnesses for Automated Wiring Masaki Yoneda, 1 Takayuki Okatani 1 and Koichiro Deguchi 1 This paper presents a method for recognizing the pose of a wire harness
4. C i k = 2 k-means C 1 i, C 2 i 5. C i x i p [ f(θ i ; x) = (2π) p 2 Vi 1 2 exp (x µ ] i) t V 1 i (x µ i ) 2 BIC BIC = 2 log L( ˆθ i ; x i C i ) + q
![4. C i k = 2 k-means C 1 i, C 2 i 5. C i x i p [ f(θ i ; x) = (2π) p 2 Vi 1 2 exp (x µ ] i) t V 1 i (x µ i ) 2 BIC BIC = 2 log L( ˆθ i ; x i C i ) + q](/thumbs/92/110416683.jpg "4. C i k = 2 k-means C 1 i, C 2 i 5. C i x i p [ f(θ i ; x) = (2π) p 2 Vi 1 2 exp (x µ ] i) t V 1 i (x µ i ) 2 BIC BIC = 2 log L( ˆθ i ; x i C i ) + q")
x-means 1 2 2 x-means, x-means k-means Bayesian Information Criterion BIC Watershed x-means Moving Object Extraction Using the Number of Clusters Determined by X-means Clustering Naoki Kubo, 1 Kousuke
スライド 1
知能制御システム学 画像追跡 (1) 特徴点の検出と追跡 東北大学大学院情報科学研究科鏡慎吾 swk(at)ic.is.tohoku.ac.jp 2008.07.07 今日の内容 前回までの基本的な画像処理の例を踏まえて, ビジュアルサーボシステムの構成要素となる画像追跡の代表的手法を概説する 画像上の ある点 の追跡 オプティカルフローの拘束式 追跡しやすい点 (Harris オペレータ ) Lucas-Kanade
IPSJ SIG Technical Report Vol.2012-CG-149 No.13 Vol.2012-CVIM-184 No /12/4 3 1,a) ( ) DB 3D DB 2D,,,, PnP(Perspective n-point), Ransa
3,a) 3 3 ( ) DB 3D DB 2D,,,, PnP(Perspective n-point), Ransac. DB [] [2] 3 DB Web Web DB Web NTT NTT Media Intelligence Laboratories, - Hikarinooka Yokosuka-Shi, Kanagawa 239-0847 Japan a) yabushita.hiroko@lab.ntt.co.jp
SICE東北支部研究集会資料(2013年)
280 (2013.5.29) 280-4 SURF A Study of SURF Algorithm using Edge Image and Color Information Yoshihiro Sasaki, Syunichi Konno, Yoshitaka Tsunekawa * *Iwate University : SURF (Speeded Up Robust Features)
1 (n = 52, 386) DL (n = 52, 386) DL DL [4] Dynamic Time Warping(DTW ) [5] Altmetrics Gunther [
![1 (n = 52, 386) DL (n = 52, 386) DL DL [4] Dynamic Time Warping(DTW ) [5] Altmetrics Gunther [](/thumbs/91/106046551.jpg "1 (n = 52, 386) DL (n = 52, 386) DL DL [4] Dynamic Time Warping(DTW ) [5] Altmetrics Gunther [")
DEIM Forum 2014 C5-6 191 0065 6 6 191 0065 6 6 432 8011 3 5 1 E-mail: {sugiyama-iori@ed., ishikawa-hiroshi@}tmu.ac.jp, endo-masaki@ed.tmu.ac.jp, yokoyama@inf.shizuoka.ac.jp (bibliometrics) h-index Dynamic
[1] SBS [2] SBS Random Forests[3] Random Forests ii
![[1] SBS [2] SBS Random Forests[3] Random Forests ii](/thumbs/74/70333192.jpg "[1] SBS [2] SBS Random Forests[3] Random Forests ii")
Random Forests 2013 3 A Graduation Thesis of College of Engineering, Chubu University Proposal of an efficient feature selection using the contribution rate of Random Forests Katsuya Shimazaki [1] SBS
Microsoft PowerPoint - pr_12_template-bs.pptx
12 回パターン検出と画像特徴 テンプレートマッチング 領域分割 画像特徴 テンプレート マッチング 1 テンプレートマッチング ( 図形 画像などの ) 型照合 Template Matching テンプレートと呼ばれる小さな一部の画像領域と同じパターンが画像全体の中に存在するかどうかを調べる方法 画像内にある対象物体の位置検出 物体数のカウント 物体移動の検出などに使われる テンプレートマッチングの計算
IPSJ SIG Technical Report iphone iphone,,., OpenGl ES 2.0 GLSL(OpenGL Shading Language), iphone GPGPU(General-Purpose Computing on Graphics Proc
iphone 1 1 1 iphone,,., OpenGl ES 2.0 GLSL(OpenGL Shading Language), iphone GPGPU(General-Purpose Computing on Graphics Processing Unit)., AR Realtime Natural Feature Tracking Library for iphone Makoto
untitled
K-Means 1 5 2 K-Means 7 2.1 K-Means.............................. 7 2.2 K-Means.......................... 8 2.3................... 9 3 K-Means 11 3.1.................................. 11 3.2..................................
Real AdaBoost HOG 2009 3 A Graduation Thesis of College of Engineering, Chubu University Efficient Reducing Method of HOG Features for Human Detection based on Real AdaBoost Chika Matsushima ITS Graphics
IPSJ SIG Technical Report Vol.2014-DPS-158 No.27 Vol.2014-CSEC-64 No /3/6 1,a) 2,b) 3,c) 1,d) 3 Cappelli Bazen Cappelli Bazen Cappelli 1.,,.,.,
1,a),b) 3,c) 1,d) 3 Cappelli Bazen Cappelli Bazen Cappelli 1.,,,,,.,,,,.,,.,,,,.,, 1 Department of Electrical Electronic and Communication Engineering Faculty of Science and Engineering Chuo University
Vol.-CVIM-7 No.7 /3/8 NLPCA kernel PCA KPCA 4),) NLPCA KPCA NLPCA KPCA principle curve principle surface KPCA ) ),),6),8),),3) ) Jacobian KPCA PCA ) P
Vol.-CVIM-7 No.7 /3/8 RANSAC M Subspace fitting via robust Jacobian kernel PCA Jun Fujiki and Shotaro Akaho The subspace fitting method based on the original kernel principle component analysis (PCA),
IPSJ SIG Technical Report Taubin Ellipse Fitting by Hyperaccurate Least Squares Yuuki Iwamoto, 1 Prasanna Rangarajan 2 and Kenichi Kanatani
1 2 1 2 Taubin Ellipse Fitting by Hyperaccurate Least Squares Yuuki Iwamoto, 1 Prasanna Rangarajan 2 and Kenichi Kanatani 1 This paper presents a new method for fitting an ellipse to a point sequence extracted
4d_06.dvi
Learning and Recognition of Time-Series Data Based on Self-Organizing Incremental Neural Network Shogo OKADA and Osamu HASEGAWA Self-Organizing Incremental Neural Network (SOINN) DP [12] DP SOINN HMM (Hidden
RANSAC RANSAC Amerini [8] RANSAC LO-RANSAC(Locally Optimized RANSAC)[9] LO-RANSAC 2.2 SIFT SIFT SIFT 128 SIFT SIFT SIFT SIFT p i p j d ij SIF
![RANSAC RANSAC Amerini [8] RANSAC LO-RANSAC(Locally Optimized RANSAC)[9] LO-RANSAC 2.2 SIFT SIFT SIFT 128 SIFT SIFT SIFT SIFT p i p j d ij SIF](/thumbs/86/93213089.jpg "RANSAC RANSAC Amerini [8] RANSAC LO-RANSAC(Locally Optimized RANSAC)[9] LO-RANSAC 2.2 SIFT SIFT SIFT 128 SIFT SIFT SIFT SIFT p i p j d ij SIF")
RANSAC 1 1 1 2 1 SIFT LO-RANSAC Improving the method detecting partially duplicated regions using RANSAC Kenji Kawashima 1 Tetsuya Matsumoto 1 Hiroaki Kudo 1 Yoshinori Takeuchi 2 Noboru Ohnishi 1 Abstract:
スライド 1
知能制御システム学 画像処理の基礎 (2) OpenCV による基本的な例 東北大学大学院情報科学研究科鏡慎吾 swk(at)ic.is.tohoku.ac.jp 2009.06.30 局所処理の例 空間フィルタリング 注目点の近傍 ( 典型的には 3x3 画素,5x5 画素,... など ) の画素値から, 出力 G x,y を定める { F i,j }, (i, j) Neighbor(x,y)
, ( ξ/) ξ(x), ( ξ/) x = x 1,. ξ ξ ( ξ, u) = 0. M LS ξ ξ (6) u,, u M LS 3).,.. ξ x ξ = ξ(x),, 1. J = (ξ ξ, V [ξ ] 1 (ξ ξ )) (7) ( ξ, u) = 0, = 1,..., N
![, ( ξ/) ξ(x), ( ξ/) x = x 1,. ξ ξ ( ξ, u) = 0. M LS ξ ξ (6) u,, u M LS 3).,.. ξ x ξ = ξ(x),, 1. J = (ξ ξ, V [ξ ] 1 (ξ ξ )) (7) ( ξ, u) = 0, = 1,..., N](/thumbs/80/81723780.jpg ", ( ξ/) ξ(x), ( ξ/) x = x 1,. ξ ξ ( ξ, u) = 0. M LS ξ ξ (6) u,, u M LS 3).,.. ξ x ξ = ξ(x),, 1. J = (ξ ξ, V [ξ ] 1 (ξ ξ )) (7) ( ξ, u) = 0, = 1,..., N")
1,,.,.. Maximum Likelihood Estimation for Geometric Fitting Yasuyuki Sugaya 1 Geometric fitting, the problem which estimates a geometric model of a scene from extracted image data, is one of the most fundamental
Microsoft PowerPoint - SSII_harada pptx
The state of the world The gathered data The processed data w d r I( W; D) I( W; R) The data processing theorem states that data processing can only destroy information. David J.C. MacKay. Information
å‰Łçı—訋çfl»æ³Łã†¨ã…Łã‡£ã…œã…−ã……ã…†æŁ°, ㆚ㆮ2æ¬¡è©Łä¾¡å‹ƒå›²ã•† ㅋㅪㅜã…−ã……ã…†æŁ°å‹Šã†«ã‡‹ã‡‰é•£ã†®ç¢ºç”⁄訋箊
, 2 August 28 (Fri), 2016 August 28 (Fri), 2016 1 / 64 Outline 1 2 3 2 4 2 5 6 August 28 (Fri), 2016 2 / 64 fibonacci Lucas 2 August 28 (Fri), 2016 3 / 64 Dynamic Programming R.Bellman Bellman Continuum
す 局所領域 ωk において 線形変換に用いる係数 (ak 画素の係数 (ak bk ) を算出し 入力画像の信号成分を bk ) は次式のコスト関数 E を最小化するように最適化 有さない画素に対して 式 (2) より画素値を算出する される これにより 低解像度な画像から補間によるアップサ E(
IR E-mail: hf@cs.chubu.ac.jp Abstract IR RGB ( ) IR IR IR RGB RGB PSNR 1 Time-Of- Flight(TOF)[1] Kinect [2] TOF LED TOF [3] [6] [4][5] 2 [6] RGB ( ) Infrared(IR) IR 2 2.1 1 す 局所領域 ωk において 線形変換に用いる係数 (ak
IS1-09 第 回画像センシングシンポジウム, 横浜,14 年 6 月 2 Hough Forest Hough Forest[6] Random Forest( [5]) Random Forest Hough Forest Hough Forest 2.1 Hough Forest 1 2.2
![IS1-09 第 回画像センシングシンポジウム, 横浜,14 年 6 月 2 Hough Forest Hough Forest[6] Random Forest( [5]) Random Forest Hough Forest Hough Forest 2.1 Hough Forest 1 2.2](/thumbs/91/106196100.jpg "IS1-09 第 回画像センシングシンポジウム, 横浜,14 年 6 月 2 Hough Forest Hough Forest[6] Random Forest( [5]) Random Forest Hough Forest Hough Forest 2.1 Hough Forest 1 2.2")
IS1-09 第 回画像センシングシンポジウム, 横浜,14 年 6 月 MI-Hough Forest () E-mail: ym@vision.cs.chubu.ac.jphf@cs.chubu.ac.jp Abstract Hough Forest Random Forest MI-Hough Forest Multiple Instance Learning Bag Hough Forest
1 -- 9 -- 6 6--1 (DFT) 009 DFT: Discrete Fourier Transform 6--1--1 N x[n] DFT N 1 X[k] = x[n]wn kn, k = 0, 1,, N 1 (6 ) n=0 1) W N = e j π N W N twidd
![1 -- 9 -- 6 6--1 (DFT) 009 DFT: Discrete Fourier Transform 6--1--1 N x[n] DFT N 1 X[k] = x[n]wn kn, k = 0, 1,, N 1 (6 ) n=0 1) W N = e j π N W N twidd](/thumbs/42/23059160.jpg "1 -- 9 -- 6 6--1 (DFT) 009 DFT: Discrete Fourier Transform 6--1--1 N x[n] DFT N 1 X[k] = x[n]wn kn, k = 0, 1,, N 1 (6 ) n=0 1) W N = e j π N W N twidd")
1 -- 9 6 009 (DFT) 6-1 DFT 6- DFT FFT 6-3 DFT 6-4 6-5 c 011 1/(0) 1 -- 9 -- 6 6--1 (DFT) 009 DFT: Discrete Fourier Transform 6--1--1 N x[n] DFT N 1 X[k] = x[n]wn kn, k = 0, 1,, N 1 (6 ) n=0 1) W N = e
IPSJ SIG Technical Report 1,a) 1,b) 1,c) 1,d) 2,e) 2,f) 2,g) 1. [1] [2] 2 [3] Osaka Prefecture University 1 1, Gakuencho, Naka, Sakai,
![IPSJ SIG Technical Report 1,a) 1,b) 1,c) 1,d) 2,e) 2,f) 2,g) 1. [1] [2] 2 [3] Osaka Prefecture University 1 1, Gakuencho, Naka, Sakai,](/thumbs/94/119099134.jpg "IPSJ SIG Technical Report 1,a) 1,b) 1,c) 1,d) 2,e) 2,f) 2,g) 1. [1] [2] 2 [3] Osaka Prefecture University 1 1, Gakuencho, Naka, Sakai,")
1,a) 1,b) 1,c) 1,d) 2,e) 2,f) 2,g) 1. [1] [2] 2 [3] 1 599 8531 1 1 Osaka Prefecture University 1 1, Gakuencho, Naka, Sakai, Osaka 599 8531, Japan 2 565 0871 Osaka University 1 1, Yamadaoka, Suita, Osaka
(MIRU2010) Geometric Context Randomized Trees Geometric Context Rand
(MIRU2010) 2010 7 Geometric Context Randomized Trees 487-8501 1200 E-mail: {fukuta,ky}@vision.cs.chubu.ac.jp, hf@cs.chubu.ac.jp Geometric Context Randomized Trees 10 3, Geometric Context, Abstract Image
(MIRU2009) cuboid cuboid SURF 6 85% Web. Web Abstract Extracting Spatio-te
(MIRU2009) 2009 7 182 8585 1 5 1 E-mail: noguchi-a@mm.cs.uec.ac.jp, yanai@cs.uec.ac.jp cuboid cuboid SURF 6 85% Web. Web Abstract Extracting Spatio-temporal Local Features Considering Consecutiveness of
1 12 ( )150 ( ( ) ) x M x 0 1 M 2 5x 2 + 4x + 3 x 2 1 M x M 2 1 M x (x + 1) 2 (1) x 2 + x + 1 M (2) 1 3 M (3) x 4 +
( )5 ( ( ) ) 4 6 7 9 M M 5 + 4 + M + M M + ( + ) () + + M () M () 4 + + M a b y = a + b a > () a b () y V a () V a b V n f() = n k= k k () < f() = log( ) t dt log () n+ (i) dt t (n + ) (ii) < t dt n+ n
http://www.ike-dyn.ritsumei.ac.jp/ hyoo/wave.html 1 1, 5 3 1.1 1..................................... 3 1.2 5.1................................... 4 1.3.......................... 5 1.4 5.2, 5.3....................
Microsoft PowerPoint - descriptor.ppt [互換モード]
![Microsoft PowerPoint - descriptor.ppt [互換モード]](/thumbs/94/118836169.jpg "Microsoft PowerPoint - descriptor.ppt [互換モード]")
1 局所特徴記述子を用いたバイオメトリクス認証の検討 A Study on Biometric Authentication Using Local Feature Descriptor 伊藤康一, 青山章一郎, 青木孝文 Koichi Ito, Shoichiro Aoyama and Takafumi Aoki 東北大学大学院情報科学研究科 Graduate School of Information
main.dvi
PC 1 1 [1][2] [3][4] ( ) GPU(Graphics Processing Unit) GPU PC GPU PC ( 2 GPU ) GPU Harris Corner Detector[5] CPU ( ) ( ) CPU GPU 2 3 GPU 4 5 6 7 1 toyohiro@isc.kyutech.ac.jp 45 2 ( ) CPU ( ) ( ) () 2.1
2007/8 Vol. J90 D No. 8 Stauffer [7] 2 2 I 1 I 2 2 (I 1(x),I 2(x)) 2 [13] I 2 = CI 1 (C >0) (I 1,I 2) (I 1,I 2) Field Monitoring Server
![2007/8 Vol. J90 D No. 8 Stauffer [7] 2 2 I 1 I 2 2 (I 1(x),I 2(x)) 2 [13] I 2 = CI 1 (C >0) (I 1,I 2) (I 1,I 2) Field Monitoring Server](/thumbs/91/107223763.jpg "2007/8 Vol. J90 D No. 8 Stauffer [7] 2 2 I 1 I 2 2 (I 1(x),I 2(x)) 2 [13] I 2 = CI 1 (C >0) (I 1,I 2) (I 1,I 2) Field Monitoring Server")
a) Change Detection Using Joint Intensity Histogram Yasuyo KITA a) 2 (0 255) (I 1 (x),i 2 (x)) I 2 = CI 1 (C>0) (I 1,I 2 ) (I 1,I 2 ) 2 1. [1] 2 [2] [3] [5] [6] [8] Intelligent Systems Research Institute,
On the Limited Sample Effect of the Optimum Classifier by Bayesian Approach he Case of Independent Sample Size for Each Class Xuexian HA, etsushi WAKA
Journal Article / 学術雑誌論文 ベイズアプローチによる最適識別系の有限 標本効果に関する考察 : 学習標本の大きさ がクラス間で異なる場合 (< 論文小特集 > パ ターン認識のための学習 : 基礎と応用 On the limited sample effect of bayesian approach : the case of each class 韓, 雪仙 ; 若林, 哲史
2008 : 80725872 1 2 2 3 2.1.......................................... 3 2.2....................................... 3 2.3......................................... 4 2.4 ()..................................
Run-Based Trieから構成される 決定木の枝刈り法
Run-Based Trie 2 2 25 6 Run-Based Trie Simple Search Run-Based Trie Network A Network B Packet Router Packet Filtering Policy Rule Network A, K Network B Network C, D Action Permit Deny Permit Network
2.2 (a) = 1, M = 9, p i 1 = p i = p i+1 = 0 (b) = 1, M = 9, p i 1 = 0, p i = 1, p i+1 = 1 1: M 2 M 2 w i [j] w i [j] = 1 j= w i w i = (w i [ ],, w i [
![2.2 (a) = 1, M = 9, p i 1 = p i = p i+1 = 0 (b) = 1, M = 9, p i 1 = 0, p i = 1, p i+1 = 1 1: M 2 M 2 w i [j] w i [j] = 1 j= w i w i = (w i [ ],, w i [](/thumbs/93/111684107.jpg "2.2 (a) = 1, M = 9, p i 1 = p i = p i+1 = 0 (b) = 1, M = 9, p i 1 = 0, p i = 1, p i+1 = 1 1: M 2 M 2 w i [j] w i [j] = 1 j= w i w i = (w i [ ],, w i [")
RI-002 Encoding-oriented video generation algorithm based on control with high temporal resolution Yukihiro BANDOH, Seishi TAKAMURA, Atsushi SHIMIZU 1 1T / CMOS [1] 4K (4096 2160 /) 900 Hz 50Hz,60Hz 240Hz
h(n) x(n) s(n) S (ω) = H(ω)X(ω) (5 1) H(ω) H(ω) = F[h(n)] (5 2) F X(ω) x(n) X(ω) = F[x(n)] (5 3) S (ω) s(n) S (ω) = F[s(n)] (5
![h(n) x(n) s(n) S (ω) = H(ω)X(ω) (5 1) H(ω) H(ω) = F[h(n)] (5 2) F X(ω) x(n) X(ω) = F[x(n)] (5 3) S (ω) s(n) S (ω) = F[s(n)] (5](/thumbs/91/106545489.jpg "h(n) x(n) s(n) S (ω) = H(ω)X(ω) (5 1) H(ω) H(ω) = F[h(n)] (5 2) F X(ω) x(n) X(ω) = F[x(n)] (5 3) S (ω) s(n) S (ω) = F[s(n)] (5")
1 -- 5 5 2011 2 1940 N. Wiener FFT 5-1 5-2 Norbert Wiener 1894 1912 MIT c 2011 1/(12) 1 -- 5 -- 5 5--1 2008 3 h(n) x(n) s(n) S (ω) = H(ω)X(ω) (5 1) H(ω) H(ω) = F[h(n)] (5 2) F X(ω) x(n) X(ω) = F[x(n)]
Design of highly accurate formulas for numerical integration in weighted Hardy spaces with the aid of potential theory 1 Ken ichiro Tanaka 1 Ω R m F I = F (t) dt (1.1) Ω m m 1 m = 1 1 Newton-Cotes Gauss
() x + y + y + x dy dx = 0 () dy + xy = x dx y + x y ( 5) ( s55906) 0.7. (). 5 (). ( 6) ( s6590) 0.8 m n. 0.9 n n A. ( 6) ( s6590) f A (λ) = det(a λi)
0. A A = 4 IC () det A () A () x + y + z = x y z X Y Z = A x y z ( 5) ( s5590) 0. a + b + c b c () a a + b + c c a b a + b + c 0 a b c () a 0 c b b c 0 a c b a 0 0. A A = 7 5 4 5 0 ( 5) ( s5590) () A ()
カルマンフィルターによるベータ推定（ ）
β TOPIX 1 22 β β smoothness priors (the Capital Asset Pricing Model, CAPM) CAPM 1 β β β β smoothness priors :,,. E-mail: koiti@ism.ac.jp., 104 1 TOPIX β Z i = β i Z m + α i (1) Z i Z m α i α i β i (the
2.2 h h l L h L = l cot h (1) (1) L l L l l = L tan h (2) (2) L l 2 l 3 h 2.3 a h a h (a, h)
1 16 10 5 1 2 2.1 a a a 1 1 1 2.2 h h l L h L = l cot h (1) (1) L l L l l = L tan h (2) (2) L l 2 l 3 h 2.3 a h a h (a, h) 4 2 3 4 2 5 2.4 x y (x,y) l a x = l cot h cos a, (3) y = l cot h sin a (4) h a
ばらつき抑制のための確率最適制御
( ) http://wwwhayanuemnagoya-uacjp/ fujimoto/ 2011 3 9 11 ( ) 2011/03/09-11 1 / 46 Outline 1 2 3 4 5 ( ) 2011/03/09-11 2 / 46 Outline 1 2 3 4 5 ( ) 2011/03/09-11 3 / 46 (1/2) r + Controller - u Plant y
2014/3 Vol. J97 D No. 3 Recognition-based segmentation [7] 1 DP 1 Conditional random field; CRF [8] [10] CRF / OCR 2 2 2 2 OCR 2 2 2 2. 2 2 2 [11], [1
![2014/3 Vol. J97 D No. 3 Recognition-based segmentation [7] 1 DP 1 Conditional random field; CRF [8] [10] CRF / OCR 2 2 2 2 OCR 2 2 2 2. 2 2 2 [11], [1](/thumbs/42/22940740.jpg "2014/3 Vol. J97 D No. 3 Recognition-based segmentation [7] 1 DP 1 Conditional random field; CRF [8] [10] CRF / OCR 2 2 2 2 OCR 2 2 2 2. 2 2 2 [11], [1")
2, a) Scene Character Extraction by an Optimal Two-Dimensional Segmentation Hiroaki TAKEBE, a) and Seiichi UCHIDA / 2 2 2 2 2 2 1. FUJITSU LABORATORIES LTD., 4 1 1 Kamikodanaka, Nakahara-ku, Kawasaki-shi,
一般社団法人電子情報通信学会 THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGIN
一般社団法人電子情報通信学会 THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS 信学技報 IEICE Technical Report PRMU2017-36,SP2017-12(2017-06)
a) Extraction of Similarities and Differences in Human Behavior Using Singular Value Decomposition Kenichi MISHIMA, Sayaka KANATA, Hiroaki NAKANISHI a
a) Extraction of Similarities and Differences in Human Behavior Using Singular Value Decomposition Kenichi MISHIMA, Sayaka KANATA, Hiroaki NAKANISHI a), Tetsuo SAWARAGI, and Yukio HORIGUCHI 1. Johansson
it-ken_open.key
深層学習技術の進展 ImageNet Classification 画像認識 音声認識 自然言語処理 機械翻訳 深層学習技術は これらの分野において 特に圧倒的な強みを見せている Figure (Left) Eight ILSVRC-2010 test Deep images and the cited4: from: ``ImageNet Classification with Networks et
第5章 偏微分方程式の境界値問題
October 5, 2018 1 / 113 4 ( ) 2 / 113 Poisson 5.1 Poisson ( A.7.1) Poisson Poisson 1 (A.6 ) Γ p p N u D Γ D b 5.1.1: = Γ D Γ N 3 / 113 Poisson 5.1.1 d {2, 3} Lipschitz (A.5 ) Γ D Γ N = \ Γ D Γ p Γ N Γ
& 3 3 ' ' (., (Pixel), (Light Intensity) (Random Variable). (Joint Probability). V., V = {,,, V }. i x i x = (x, x,, x V ) T. x i i (State Variable),
.... Deeping and Expansion of Large-Scale Random Fields and Probabilistic Image Processing Kazuyuki Tanaka The mathematical frameworks of probabilistic image processing are formulated by means of Markov
( ) sin 1 x, cos 1 x, tan 1 x sin x, cos x, tan x, arcsin x, arccos x, arctan x. π 2 sin 1 x π 2, 0 cos 1 x π, π 2 < tan 1 x < π 2 1 (1) (
6 20 ( ) sin, cos, tan sin, cos, tan, arcsin, arccos, arctan. π 2 sin π 2, 0 cos π, π 2 < tan < π 2 () ( 2 2 lim 2 ( 2 ) ) 2 = 3 sin (2) lim 5 0 = 2 2 0 0 2 2 3 3 4 5 5 2 5 6 3 5 7 4 5 8 4 9 3 4 a 3 b
pp d 2 * Hz Hz 3 10 db Wind-induced noise, Noise reduction, Microphone array, Beamforming 1
72 12 2016 pp. 739 748 739 43.60.+d 2 * 1 2 2 3 2 125 Hz 0.3 0.8 2 125 Hz 3 10 db Wind-induced noise, Noise reduction, Microphone array, Beamforming 1. 1.1 PSS [1] [2 4] 2 Wind-induced noise reduction
3 2 2 (1) (2) (3) (4) 4 4 AdaBoost 2. [11] Onishi&Yoda [8] Iwashita&Stoica [5] 4 [3] 3. 3 (1) (2) (3)
![3 2 2 (1) (2) (3) (4) 4 4 AdaBoost 2. [11] Onishi&Yoda [8] Iwashita&Stoica [5] 4 [3] 3. 3 (1) (2) (3)](/thumbs/68/59204737.jpg "3 2 2 (1) (2) (3) (4) 4 4 AdaBoost 2. [11] Onishi&Yoda [8] Iwashita&Stoica [5] 4 [3] 3. 3 (1) (2) (3)")
(MIRU2012) 2012 8 820-8502 680-4 E-mail: {d kouno,shimada,endo}@pluto.ai.kyutech.ac.jp (1) (2) (3) (4) 4 AdaBoost 1. Kanade [6] CLAFIC [12] EigenFace [10] 1 1 2 1 [7] 3 2 2 (1) (2) (3) (4) 4 4 AdaBoost
ds2.dvi
1 Fourier 2 : Fourier s(t) Fourier S(!) = s(t) = 1 s(t)e j!t dt (1) S(!)e j!t d! (2) 1 1 s(t) S(!) S(!) =S Λ (!) Λ js T (!)j 2 P (!) = lim T!1 T S T (!) = T=2 T=2 (3) s(t)e j!t dt (4) T P (!) Fourier P
miru2006_cr.dvi
y;yy yy y y;yy yy y;yy y 630 0192 8916 5 yy NEC 630 0101 8916 47 E-mail: yftomoka-s,sei-i,kanbara,yokoyag@is.naist.jp, yyiketani@cp, n-nakajima@ay.jp.nec.com (structure from motion), structure from motion,,,
THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS TECHNICAL REPORT OF IEICE. UWB UWB
THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS TECHNICAL REPORT OF IEICE. UWB -1 E-mail: seki@aso.cce.i.koto-u.ac.jp UWB SEABED SEABED SEABED,,, SEABED Application of fast imaging
0A_SeibutsuJyoho-RF.ppt
A ON-Center OFF-Center DeAngelis, Ohzawa, Freeman 1995 Nobel Prize 1981: Physiology and Medicine D.H. Hubel and T.N. Wiesel T.N. Wiesel D.H. Hubel V1/V2: (spikes) Display? Amplifiers and Filters V1 - simple
(4) ω t(x) = 1 ω min Ω ( (I C (y))) min 0 < ω < C A C = 1 (5) ω (5) t transmission map tmap 1 4(a) 2. 3 2. 2 t 4(a) t tmap RGB 2 (a) RGB (A), (B), (C)
(MIRU2011) 2011 7 890 0065 1 21 40 105-6691 1 1 1 731 3194 3 4 1 338 8570 255 346 8524 1836 1 E-mail: {fukumoto,kawasaki}@ibe.kagoshima-u.ac.jp, ryo-f@hiroshima-cu.ac.jp, fukuda@cv.ics.saitama-u.ac.jp,
(1.2) T D = 0 T = D = 30 kn 1.2 (1.4) 2F W = 0 F = W/2 = 300 kn/2 = 150 kn 1.3 (1.9) R = W 1 + W 2 = = 1100 N. (1.9) W 2 b W 1 a = 0
1 1 1.1 1.) T D = T = D = kn 1. 1.4) F W = F = W/ = kn/ = 15 kn 1. 1.9) R = W 1 + W = 6 + 5 = 11 N. 1.9) W b W 1 a = a = W /W 1 )b = 5/6) = 5 cm 1.4 AB AC P 1, P x, y x, y y x 1.4.) P sin 6 + P 1 sin 45
seminar0220a.dvi
1 Hi-Stat 2 16 2 20 16:30-18:00 2 2 217 1 COE 4 COE RA E-MAIL: ged0104@srv.cc.hit-u.ac.jp 2004 2 25 S-PLUS S-PLUS S-PLUS S-code 2 [8] [8] [8] 1 2 ARFIMA(p, d, q) FI(d) φ(l)(1 L) d x t = θ(l)ε t ({ε t }
IPSJ SIG Technical Report Vol.2012-CVIM-182 No /5/ RGB [1], [2], [3], [4], [5] [6], [7], [8], [9] 1 (MSFA: Multi-Spectrum Filt
![IPSJ SIG Technical Report Vol.2012-CVIM-182 No /5/ RGB [1], [2], [3], [4], [5] [6], [7], [8], [9] 1 (MSFA: Multi-Spectrum Filt](/thumbs/104/162109798.jpg "IPSJ SIG Technical Report Vol.2012-CVIM-182 No /5/ RGB [1], [2], [3], [4], [5] [6], [7], [8], [9] 1 (MSFA: Multi-Spectrum Filt")
1 1 1 1 1. 4 3 RGB [1], [2], [3], [4], [5] [6], [7], [8], [9] 1 (MSFA: Multi-Spectrum Filter Array) 1 [8], [9] RGB 1 Tokyo Institute of Technology 1 [10], [11], [12], [13], [14] [15] Parmar Wiener RGB
DEIM Forum 2019 A7-1 Flexible Distance-based Hashing mori
DEIM Forum 2019 A7-1 Flexible Distance-based Hashing 731 3194 E-mail: mc66023@e.hiroshima-cu.ac.jp,{wakaba,s naga,inagi,yoko}@hiroshima-cu.ac.jp, morikei18@gmail.com Flexible Distance-based Hashing(FDH)
THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS TECHNICAL REPORT OF IEICE. TV A310
THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS TECHNICAL REPORT OF IEICE. TV 367 0035 1011 A310 E-mail kawamura@suou.waseda.jp Total Variation Total Variation Total Variation Abstract
1 : ( ) ( ) ( ) ( ) ( ) etc (SCA)
START: 17th Symp. Auto. Decentr. Sys., Jan. 28, 2005 Symplectic cellular automata as a test-bed for research on the emergence of natural systems 1 : ( ) ( ) ( ) ( ) ( ) etc (SCA) 2 SCA 2.0 CA ( ) E.g.
proc.dvi
The Great Buddha Project Λ1 Λ2 Λ2 Λ2 Λ3 Λ2 Λ1 Λ2 Λ3 ( (VR) VR (1) (2) (3) modeling-from-reality (MFR) The Great Buddha Project Digital archive of large-scale cultural heritage Ryo Kurazume Λ1 Ko Nishino
paper.dvi
23 Study on character extraction from a picture using a gradient-based feature 1120227 2012 3 1 Google Street View Google Street View SIFT 3 SIFT 3 y -80 80-50 30 SIFT i Abstract Study on character extraction
,,, 2 ( ), $[2, 4]$, $[21, 25]$, $V$,, 31, 2, $V$, $V$ $V$, 2, (b) $-$,,, (1) : (2) : (3) : $r$ $R$ $r/r$, (4) : 3
![,,, 2 ( ), $[2, 4]$, $[21, 25]$, $V$,, 31, 2, $V$, $V$ $V$, 2, (b) $-$,,, (1) : (2) : (3) : $r$ $R$ $r/r$, (4) : 3](/thumbs/91/104771717.jpg ",,, 2 ( ), $[2, 4]$, $[21, 25]$, $V$,, 31, 2, $V$, $V$ $V$, 2, (b) $-$,,, (1) : (2) : (3) : $r$ $R$ $r/r$, (4) : 3")
1084 1999 124-134 124 3 1 (SUGIHARA Kokichi),,,,, 1, [5, 11, 12, 13], (2, 3 ), -,,,, 2 [5], 3,, 3, 2 2, -, 3,, 1,, 3 2,,, 3 $R$ ( ), $R$ $R$ $V$, $V$ $R$,,,, 3 2 125 1 3,,, 2 ( ), $[2, 4]$, $[21, 25]$,
1 (Berry,1975) 2-6 p (S πr 2 )p πr 2 p 2πRγ p p = 2γ R (2.5).1-1 : : : : ( ).2 α, β α, β () X S = X X α X β (.1) 1 2
2005 9/8-11 2 2.2 ( 2-5) γ ( ) γ cos θ 2πr πρhr 2 g h = 2γ cos θ ρgr (2.1) γ = ρgrh (2.2) 2 cos θ θ cos θ = 1 (2.2) γ = 1 ρgrh (2.) 2 2. p p ρgh p ( ) p p = p ρgh (2.) h p p = 2γ r 1 1 (Berry,1975) 2-6
Vol. 44 No. SIG 9(CVIM 7) ) 2) 1) 1 2) 3 7) 1) 2) 3 3) 4) 5) (a) (d) (g) (b) (e) (h) No Convergence? End (f) (c) Yes * ** * ** 1
Vol. 44 No. SIG 9(CVIM 7) July 2003, Robby T. Tan, 1 Estimating Illumination Position, Color and Surface Reflectance Properties from a Single Image Kenji Hara,, Robby T. Tan, Ko Nishino, Atsushi Nakazawa,
18 2 20 W/C W/C W/C 4-4-1 0.05 1.0 1000 1. 1 1.1 1 1.2 3 2. 4 2.1 4 (1) 4 (2) 4 2.2 5 (1) 5 (2) 5 2.3 7 3. 8 3.1 8 3.2 ( ) 11 3.3 11 (1) 12 (2) 12 4. 14 4.1 14 4.2 14 (1) 15 (2) 16 (3) 17 4.3 17 5. 19
(3.6 ) (4.6 ) 2. [3], [6], [12] [7] [2], [5], [11] [14] [9] [8] [10] (1) Voodoo 3 : 3 Voodoo[1] 3 ( 3D ) (2) : Voodoo 3D (3) : 3D (Welc
![(3.6 ) (4.6 ) 2. [3], [6], [12] [7] [2], [5], [11] [14] [9] [8] [10] (1) Voodoo 3 : 3 Voodoo[1] 3 ( 3D ) (2) : Voodoo 3D (3) : 3D (Welc](/thumbs/65/54056306.jpg "(3.6 ) (4.6 ) 2. [3], [6], [12] [7] [2], [5], [11] [14] [9] [8] [10] (1) Voodoo 3 : 3 Voodoo[1] 3 ( 3D ) (2) : Voodoo 3D (3) : 3D (Welc")
1,a) 1,b) Obstacle Detection from Monocular On-Vehicle Camera in units of Delaunay Triangles Abstract: An algorithm to detect obstacles by using a monocular on-vehicle video camera is developed. Since
Accuracy Improvement by Compound Discriminant Functions for Resembling Character Recognition Takashi NAKAJIMA, Tetsushi WAKABAYASHI, Fumitaka KIMURA,
Journal Article / 学 術 雑 誌 論 文 混 合 識 別 関 数 による 類 似 文 字 認 識 の 高 精 度 化 Accuracy improvement by compoun for resembling character recogn 中 嶋, 孝 ; 若 林, 哲 史 ; 木 村, 文 隆 ; 三 宅, 康 二 Nakajima, Takashi; Wakabayashi,
28 Horizontal angle correction using straight line detection in an equirectangular image
28 Horizontal angle correction using straight line detection in an equirectangular image 1170283 2017 3 1 2 i Abstract Horizontal angle correction using straight line detection in an equirectangular image
Hanbury-Brown Twiss (ver. 2.0) van Cittert - Zernike mutual coherence
Hanbury-Brown Twiss (ver. 2.) 25 4 4 1 2 2 2 2.1 van Cittert - Zernike..................................... 2 2.2 mutual coherence................................. 4 3 Hanbury-Brown Twiss ( ) 5 3.1............................................
G H J(g, τ G g G J(g, τ τ J(g 1 g, τ = J(g 1, g τj(g, τ J J(1, τ = 1 k g = ( a b c d J(g, τ = (cτ + dk G = SL (R SL (R G G α, β C α = α iθ (θ R
1 1.1 SL (R 1.1.1 SL (R H SL (R SL (R H H H = {z = x + iy C; x, y R, y > 0}, SL (R = {g M (R; dt(g = 1}, gτ = aτ + b a b g = SL (R cτ + d c d 1.1. Γ H H SL (R f(τ f(gτ G SL (R G H J(g, τ τ g G Hol f(τ
11) 13) 11),12) 13) Y c Z c Image plane Y m iy O m Z m Marker coordinate system T, d X m f O c X c Camera coordinate system 1 Coordinates and problem
1 1 1 Posture Esimation by Using 2-D Fourier Transform Yuya Ono, 1 Yoshio Iwai 1 and Hiroshi Ishiguro 1 Recently, research fields of augmented reality and robot navigation are actively investigated. Estimating
Isogai, T., Building a dynamic correlation network for fat-tailed financial asset returns, Applied Network Science (7):-24, 206,
H28. (TMU) 206 8 29 / 34 2 3 4 5 6 Isogai, T., Building a dynamic correlation network for fat-tailed financial asset returns, Applied Network Science (7):-24, 206, http://link.springer.com/article/0.007/s409-06-0008-x
(a) (b) (c) (d) 1: (a) (b) (c) (d) (a) (b) (c) 2: (a) (b) (c) 1(b) [1 10] 1 degree k n(k) walk path 4
![(a) (b) (c) (d) 1: (a) (b) (c) (d) (a) (b) (c) 2: (a) (b) (c) 1(b) [1 10] 1 degree k n(k) walk path 4](/thumbs/91/105665060.jpg "(a) (b) (c) (d) 1: (a) (b) (c) (d) (a) (b) (c) 2: (a) (b) (c) 1(b) [1 10] 1 degree k n(k) walk path 4")
1 vertex edge 1(a) 1(b) 1(c) 1(d) 2 (a) (b) (c) (d) 1: (a) (b) (c) (d) 1 2 6 1 2 6 1 2 6 3 5 3 5 3 5 4 4 (a) (b) (c) 2: (a) (b) (c) 1(b) [1 10] 1 degree k n(k) walk path 4 1: Zachary [11] [12] [13] World-Wide
2 (S, C, R, p, q, S, C, ML ) S = {s 1, s 2,..., s n } C = {c 1, c 2,..., c m } n = S m = C R = {r 1, r 2,...} r r 2 C \ p = (p r ) r R q = (q r ) r R
RF-004 Hashimoto Naoyuki Suguru Ueda Atsushi Iwasaki Yosuke Yasuda Makoto Yokoo 1 [10] ( ). ( ) 1 ( ) 3 4 3 4 = 12 deferred acceptance (DA) [3, 7] [5] ( ) NP serial dictatorship with regional quotas (SDRQ)
22_05.dvi
Vol. 1 No. 2 41 49 (July 2008) 3 1 1 3 2 1 1 3 Person-independent Monocular Tracking of Face and Facial Actions Yusuke Sugano 1 and Yoichi Sato 1 This paper presents a monocular method of tracking faces
2. 30 Visual Words TF-IDF Lowe [4] Scale-Invarient Feature Transform (SIFT) Bay [1] Speeded Up Robust Features (SURF) SIFT 128 SURF 64 Visual Words Ni
![2. 30 Visual Words TF-IDF Lowe [4] Scale-Invarient Feature Transform (SIFT) Bay [1] Speeded Up Robust Features (SURF) SIFT 128 SURF 64 Visual Words Ni](/thumbs/86/94686225.jpg "2. 30 Visual Words TF-IDF Lowe [4] Scale-Invarient Feature Transform (SIFT) Bay [1] Speeded Up Robust Features (SURF) SIFT 128 SURF 64 Visual Words Ni")
DEIM Forum 2012 B5-3 606 8510 E-mail: {zhao,ohshima,tanaka}@dl.kuis.kyoto-u.ac.jp Web, 1. Web Web TinEye 1 Google 1 http://www.tineye.com/ 1 2. 3. 4. 5. 6. 2. 30 Visual Words TF-IDF Lowe [4] Scale-Invarient
compact compact Hermann compact Hermite ( - ) Hermann Hermann ( ) compact Hermite Lagrange compact Hermite ( ) a, Σ a {0} a 3 1
014 5 4 compact compact Hermann compact Hermite ( - ) Hermann Hermann ( ) compact Hermite Lagrange compact Hermite ( ) 1 1.1. a, Σ a {0} a 3 1 (1) a = span(σ). () α, β Σ s α β := β α,β α α Σ. (3) α, β
情報処理学会研究報告 IPSJ SIG Technical Report Vol.2013-CVIM-186 No /3/15 EMD 1,a) SIFT. SIFT Bag-of-keypoints. SIFT SIFT.. Earth Mover s Distance
EMD 1,a) 1 1 1 SIFT. SIFT Bag-of-keypoints. SIFT SIFT.. Earth Mover s Distance (EMD), Bag-of-keypoints,. Bag-of-keypoints, SIFT, EMD, A method of similar image retrieval system using EMD and SIFT Hoshiga
2S III IV K A4 12:00-13:30 Cafe David 1 2 TA 1 appointment Cafe David K2-2S04-00 : C
2S III IV K200 : April 16, 2004 Version : 1.1 TA M2 TA 1 10 2 n 1 ɛ-δ 5 15 20 20 45 K2-2S04-00 : C 2S III IV K200 60 60 74 75 89 90 1 email 3 4 30 A4 12:00-13:30 Cafe David 1 2 TA 1 email appointment Cafe
2 (March 13, 2010) N Λ a = i,j=1 x i ( d (a) i,j x j ), Λ h = N i,j=1 x i ( d (h) i,j x j ) B a B h B a = N i,j=1 ν i d (a) i,j, B h = x j N i,j=1 ν i
1. A. M. Turing [18] 60 Turing A. Gierer H. Meinhardt [1] : (GM) ) a t = D a a xx µa + ρ (c a2 h + ρ 0 (0 < x < l, t > 0) h t = D h h xx νh + c ρ a 2 (0 < x < l, t > 0) a x = h x = 0 (x = 0, l) a = a(x,
24 I ( ) 1. R 3 (i) C : x 2 + y 2 1 = 0 (ii) C : y = ± 1 x 2 ( 1 x 1) (iii) C : x = cos t, y = sin t (0 t 2π) 1.1. γ : [a, b] R n ; t γ(t) = (x
![24 I ( ) 1. R 3 (i) C : x 2 + y 2 1 = 0 (ii) C : y = ± 1 x 2 ( 1 x 1) (iii) C : x = cos t, y = sin t (0 t 2π) 1.1. γ : [a, b] R n ; t γ(t) = (x](/thumbs/94/118744578.jpg "24 I ( ) 1. R 3 (i) C : x 2 + y 2 1 = 0 (ii) C : y = ± 1 x 2 ( 1 x 1) (iii) C : x = cos t, y = sin t (0 t 2π) 1.1. γ : [a, b] R n ; t γ(t) = (x")
24 I 1.1.. ( ) 1. R 3 (i) C : x 2 + y 2 1 = 0 (ii) C : y = ± 1 x 2 ( 1 x 1) (iii) C : x = cos t, y = sin t (0 t 2π) 1.1. γ : [a, b] R n ; t γ(t) = (x 1 (t), x 2 (t),, x n (t)) ( ) ( ), γ : (i) x 1 (t),
(interferometer) 1 N *3 2 ω λ k = ω/c = 2π/λ ( ) r E = A 1 e iφ1(r) e iωt + A 2 e iφ2(r) e iωt (1) φ 1 (r), φ 2 (r) r λ 2π 2 I = E 2 = A A 2 2 +
7 1 (Young) *1 *2 (interference) *1 (1802 1804) *2 2 (2005) (1993) 1 (interferometer) 1 N *3 2 ω λ k = ω/c = 2π/λ ( ) r E = A 1 e iφ1(r) e iωt + A 2 e iφ2(r) e iωt (1) φ 1 (r), φ 2 (r) r λ 2π 2 I = E 2
icde_5a_3
ICDE 2016 & WWW 2016 勉強会 Research Session 5A-3: Durable Graph Pattern Queries on Historical Graphs Konstantinos Semertzidis Evaggelia Pitoura 担当 : 楠和馬 ( 同志社大学 ) I. Introduction (1 / 2) } 背景 } 様々なドメインで時間経過につれ変化するグラフがほとんど
2 CAD : CAD 7
1 CAD 2017.6.25 2 CAD 2 3 1998 1 0 6 : CAD 7 3 CAD 2017 6 4 0 7 0.1 1............................. 7 0.2 2............................. 8 0.3 3............................ 9 0.4 4............................
Fig Measurement data combination. 2 Fig. 2. Ray vector. Fig (12) 1 2 R 1 r t 1 3 p 1,i i 2 3 Fig.2 R 2 t 2 p 2,i [u, v] T (1)(2) r R 1 R 2
![Fig Measurement data combination. 2 Fig. 2. Ray vector. Fig (12) 1 2 R 1 r t 1 3 p 1,i i 2 3 Fig.2 R 2 t 2 p 2,i [u, v] T (1)(2) r R 1 R 2](/thumbs/93/113464882.jpg "Fig Measurement data combination. 2 Fig. 2. Ray vector. Fig (12) 1 2 R 1 r t 1 3 p 1,i i 2 3 Fig.2 R 2 t 2 p 2,i [u, v] T (1)(2) r R 1 R 2")
IP 06 16 / IIS 06 32 3 3-D Environment Modeling from Images Acquired with an Omni-Directional Camera Mounted on a Mobile Robot Atsushi Yamashita, Tomoaki Harada, Ryosuke Kawanishi, Toru Kaneko (Shizuoka
微分積分 サンプルページ この本の定価 判型などは, 以下の URL からご覧いただけます. このサンプルページの内容は, 初版 1 刷発行時のものです.
微分積分 サンプルページ この本の定価 判型などは, 以下の URL からご覧いただけます. ttp://www.morikita.co.jp/books/mid/00571 このサンプルページの内容は, 初版 1 刷発行時のものです. i ii 014 10 iii [note] 1 3 iv 4 5 3 6 4 x 0 sin x x 1 5 6 z = f(x, y) 1 y = f(x)
10_08.dvi
476 67 10 2011 pp. 476 481 * 43.72.+q 1. MOS Mean Opinion Score ITU-T P.835 [1] [2] [3] Subjective and objective quality evaluation of noisereduced speech. Takeshi Yamada, Shoji Makino and Nobuhiko Kitawaki