ユーザ負担のない話者・環境適応性を実現する自然な音声対話処理技術
|
|
|
- ただきよ しんまつ
- 9 years ago
- Views:
Transcription
1
2
3 DSP
4 DSP
5
6 Julius H18 H16 H19
7
8 / (2003.3) ( ) Julian( )
9 JNAS,.,. CSRC,..,..
10 /2,800msecshort reject
11
12
13 Julius #Downloads per month Julius 4.
14
15 NAM BSSA
16 H17
17 (NAM) (NAM) NAM NAM
18 NAM
19
20 Bluetooth Bluetooth 2
21 Earphone Microphone
22 SIMO-Model Accuracy [db] FDICA-PB MS-SIMO-ICA Proposed SG Proposed SIMO-ICA-SG 1 FDICA-PB MS-SIMO-ICA Output 1 Output 1 Input Output 2 Output Direction of source 2 θ 2 [deg] Output 1 Output 2
23 SIMO-ICA IEEE/RSJ (IROS2005) Best Application Paper Award IEEE MLSP2007 Data Analysis Competition Winner on Nonlinear Separation
24 User s Speech Noise F F T X J X j ( f,τ ) ( f,τ ) Phase Compensation θ U FD- ICA Y DS ( f,τ ) User Primary Path 0 Noise Mel-Scale Filter Bank PB E j E J ( f,τ ) ( f,τ ) + Spectral Subtract Reference Path m( l,τ ) θ U - Z ( f ICA,τ ) BSS m( L,τ ) Log Transform and DCT Mel-Scale Filter Bank MFCC ( n, τ )
25 8 9 % BSSA
26
27
28 HMM Alan Black CMU) Blizzard Challenge Evaluating corpus-based speech synthesis on common data sets
29 The Blizzard Challenge Speech experts Volunteers HTS STRAIGHT Global Variance US undergraduates
30
31
32
33
34
2 1,384,000 2,000,000 1,296,211 1,793,925 38,000 54,500 27,804 43,187 41,000 60,000 31,776 49,017 8,781 18,663 25,000 35,300 3 4 5 6 1,296,211 1,793,925 27,804 43,187 1,275,648 1,753,306 29,387 43,025
VISPO /表1-4
7 2005 1,132 5,249 362 13,666 311,809 1,359 3,723 1,669 538 3,737 17,418 39,036 75,694 5,281 1,169 161,502 7,463 11,408,436 500,000 13,263 192,052 41,391 49,706 136,232 61,102 12,402,182 11,573,898 273,042
Netcommunity SYSTEM X7000 IPコードレス電話機 取扱説明書
4 5 6 7 8 9 . 4 DS 0 4 5 4 4 4 5 5 6 7 8 9 0 4 5 6 7 8 9 4 5 6 4 0 4 4 4 4 5 6 7 8 9 40 4 4 4 4 44 45 4 6 7 5 46 47 4 5 6 48 49 50 5 4 5 4 5 6 5 5 6 4 54 4 5 6 7 55 5 6 4 56 4 5 6 57 4 5 6 7 58 4
.A. D.S
1999-1- .A. D.S 1996 2001 1999-2- -3- 1 p.16 17 18 19 2-4- 1-5- 1~2 1~2 2 5 1 34 2 10 3 2.6 2.85 3.05 2.9 2.9 3.16 4 7 9 9 17 9 25 10 3 10 8 10 17 10 18 10 22 11 29-6- 1 p.1-7- p.5-8- p.9 10 12 13-9- 2
18 (1) US (2) US US US 90 (3) 2 8 1 18 108 2 2,000 3 6,000 4 33 2 17 5 2 3 1 2 8 6 7 7 2 2,000 8 1 8 19 9 10 2 2 7 11 2 12 28 1 2 11 7 1 1 1 1 1 1 3 2 3 33 2 1 3 2 3 2 16 2 8 3 28 8 3 5 13 1 14 15 1 2
2005 No ,000 8,000 7,000 6,000 5, , , ,000 4, ,000 3,000 2,000 1, , , , , ,875.0
2005No.1322 9,000 8,000 7,000 6,000 5,471.8 6,657.6 8,353.9 5,000 4,555.3 4,000 3,000 2,000 1,000 483.5 896.9 1,438.2 2,063.6 2,810.7 3,281.8 3,875.0 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
受付案内ロボットASKA
ASKA Copyright 2002 NAIST 83.5 86.3 1.1 10.1 ASKA 42.3 52.8 1.6 51.2 88.0 89.9 0.3 26.5 93.7 94.7 4.0 50.5 % % % RWCP) NEDO 30 ASKA ASKA ASKA 2001.7 NHK 2001.10 Web JNAS(306) 40kwords PTM JULIUS
.......p...{..P01-48(TF)
1 2 3 5 6 7 8 9 10 Act Plan Check Act Do Plan Check Do 11 12 13 14 INPUT OUTPUT 16 17 18 19 20 21 22 23 24 25 26 27 30 33 32 33 34 35 36 37 36 37 38 33 40 41 42 43 44 45 46 47 48 49 50 51 1. 2. 3.
2 DS SS (SS+DS) Fig. 2 Separation algorithm for motorcycle sound by combining DS and SS (SS+DS). 3. [3] DS SS 2 SS+DS 1 1 B SS SS 4. NMF 4. 1 (NMF) Y
![2 DS SS (SS+DS) Fig. 2 Separation algorithm for motorcycle sound by combining DS and SS (SS+DS). 3. [3] DS SS 2 SS+DS 1 1 B SS SS 4. NMF 4. 1 (NMF) Y](/thumbs/49/25425602.jpg "2 DS SS (SS+DS) Fig. 2 Separation algorithm for motorcycle sound by combining DS and SS (SS+DS). 3. [3] DS SS 2 SS+DS 1 1 B SS SS 4. NMF 4. 1 (NMF) Y")
a) Separation of Motorcycle Sound by Near Field Microphone Array and Nonnegative Matrix Factorization Chisaki YOSHINAGA, Nonmember, Yosuke TATEKURA a), Member, Kazuaki HAMADA, and Tetsuya KIMURA, Nonmembers
OPA277/2277/4277 (2000.1)
R OPA OPA OPA OPA OPA OPA OPA OPA OPA µ µ ± ± µ OPA ±± ±± ± µ Offset Trim Offset Trim In OPA +In -Pin DIP, SO- Output NC OPA Out A In A +In A A D Out D In D +In D Out A In A +In A A B Out B In B +In B
IEEE ax：第 6 世代の Wi-Fi テクニカル ホワイト ペーパー
IEEE 802.11ax 6 Wi-Fi 1 6 Wi-Fi IEEE 802.11ax 802.11ac IEEE 802.11ax LTE IEEE 802.11ax LAN WLAN Wi-Fi IEEE 802.11ax LAN 4K Ultra HD Internet of Things IoT 802.11ac IoT IEEE 802.11ax 3 1024 QAM 35% OFDMA
Since
Since 1885 16.4 15.6 14.2.8 39.4 13.6 21 211 212 213 214 8,88 227 186,164 181,869 184,412 197,495 21,995 11,466 11,67 15,474 16,899 22,1 1, 24, 3,54 5,686 7,518 9,615 17,467 8,88 8,37 8,288 8,29 8,319
CSR報告書2006
2006 CONTENTS 1 5 6 7 9 11 13 15 18 19 21 23 25 27 29 31 33 35 37 39 41 44 45 47 49 53 54 q w 15000 14,256 14,154 12,051 12,070 12,038 12000 9,672 10,594 9000 8,821 8,858 8,931 6000 3,848 3,004
ADC121S Bit, ksps, Diff Input, Micro Pwr Sampling ADC (jp)
ADC121S625 ADC121S625 12-Bit, 50 ksps to 200 ksps, Differential Input, Micro Power Sampling A/D Converter Literature Number: JAJSAB8 ADC121S625 12 50kSPS 200kSPS A/D ADC121S625 50kSPS 200kSPS 12 A/D 500mV
IPSJ SIG Technical Report Vol.2015-MUS-107 No /5/23 HARK-Binaural Raspberry Pi 2 1,a) ( ) HARK 2 HARK-Binaural A/D Raspberry Pi 2 1.
HARK-Binaural Raspberry Pi 2 1,a) 1 1 1 2 3 () HARK 2 HARK-Binaural A/D Raspberry Pi 2 1. [1,2] [2 5] () HARK (Honda Research Institute Japan audition for robots with Kyoto University) *1 GUI ( 1) Python
HardCopy IIデバイスのタイミング制約
7. HardCopy II H51028-2.1 Stratix II FPGA FPGA ASIC HardCopy II ASIC NRE Quartus II HardCopy Design Center HCDC Quartus II TimeQuest HardCopy II 2 DR2 TimeQuest TimeQuest FPGA ASIC FPGA ASIC Quartus II
dp0206_j
1 2 7 8 9 4 5 6 3 10 11 12 13 20 19 18 16 17 15 14 CONTROL AUDIO 2 AUDIO 1 CN2 CN1 CN20 CN15 1 2 3 4 H C E (E) 1 2 3 E C H 1 2 3 4 C 8 7 6 5 6 7 8 9 5 4 3 2 1 E A B SIGNAL/
General Purpose, Low Voltage, Rail-to-Rail Output Operational Amplifiers 324 V LM LMV321( )/LMV358( )/LMV324( ) General Purpose, Low Voltage, Rail-to-
General Purpose, Low Voltage, Rail-to-Rail Output Operational Amplifiers 324 V LM LMV321( )/LMV358( )/LMV324( ) General Purpose, Low Voltage, Rail-to-Rail Output Operational Amplifiers 358 LMV358/324 LM358/324
LMC6022 Low Power CMOS Dual Operational Amplifier (jp)
Low Power CMOS Dual Operational Amplifier Literature Number: JAJS754 CMOS CMOS (100k 5k ) 0.5mW CMOS CMOS LMC6024 100k 5k 120dB 2.5 V/ 40fA Low Power CMOS Dual Operational Amplifier 19910530 33020 23900
LT 低コスト、シャットダウン機能付き デュアルおよびトリプル300MHz 電流帰還アンプ
µ µ LT1398/LT1399 V IN A R G 00Ω CHANNEL A SELECT EN A R F 3Ω B C 97.6Ω CABLE V IN B R G 00Ω EN B R F 3Ω 97.6Ω V OUT OUTPUT (00mV/DIV) EN C V IN C 97.6Ω R G 00Ω R F 3Ω 1399 TA01 R F = R G = 30Ω f = 30MHz
00_nikke_160216
12, 1, 97,357 87,659 8, 6, 4, 2, 211 212 97,677 1,477 12,854 213 214 215 8, 6, 7,799 6,635 6,23 4,942 5,41 4,69 4, 3,12 3,261 3,346 3,572 2, 211 212 213 214 215 6, 4, 2, 3,76 12,854 4,466 4,583 3,883 4,64
JAJP.qxd
Agilent PSA LAN 217 LAN LAN PSALAN 217 802.11a/b/gLAN 6 6 12 8 8 11 CCDF 9 2 3 PSALAN 802.11b DSSS/CCK/PBCC 802.11a OFDM 802.11g OFDM/ PHY ERP 217 IEEE EVM IEEE 802.11 EVMEVM CCDF PSA 140 40 MHz 122 80
LMV851/LMV852/LMV854 8 MHz Low Power CMOS, EMI Hardened Operational Amplifi(jp)
LMV851,LMV852,LMV854 LMV851/LMV852/LMV854 8 MHz Low Power CMOS, EMI Hardened Operational Amplifiers Literature Number: JAJSAM3 LMV851/LMV852/LMV854 8MHz CMOS EMI LMV851/LMV852/LMV854 CMOS IC 40 125 LMV851/
untitled
10 log 10 W W 10 L W = 10 log 10 W 10 12 10 log 10 I I 0 I 0 =10 12 I = P2 ρc = ρcv2 L p = 10 log 10 p 2 p 0 2 = 20 log 10 p p = 20 log p 10 0 2 10 5 L 3 = 10 log 10 10 L 1 /10 +10 L 2 ( /10 ) L 1 =10
MCS-8M
4-296-435-02 (1) Volume I MCS-8M 2011 Sony Corporation m a b c I-2 ...I-6...I-7...I-8...I-8...I-9 1...I-10...I-11 1 2 3 4 5 1 2 3 4 5 2...I-17...I-19...I-21...I-23...I-26 3...I-28...I-31...I-35...I-36...I-37...I-39...I-40...I-41...I-42...I-42...I-46...I-47
IEEE ZigBee 2.4GHz 250kbps O-QPSK DSSS Bluetooth IEEE GHz 3Mbps G-FSK FHSS PC LAN IEEE b 2.4GHz 11Mbps CCK DSSS LAN LAN IEE
SMK SMK Corporation Kenji OTSUKA AV AV RF 2.4GHz ISM 2.4GHz ISM 2.4GHz RF IEEE 802.15.4 ZigBee 2.4GHz 250kbps O-QPSK DSSS Bluetooth IEEE 802.15.1 2.4GHz 3Mbps G-FSK FHSS PC LAN IEEE 802.11b 2.4GHz 11Mbps
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
7) 8) 9),10) 11) 18) 11),16) 18) 19) 20) Vocaloid 6) Vocaloid 1 VocaListener1 2 VocaListener1 3 VocaListener VocaListener1 VocaListener1 Voca
VocaListener2: 1 1 VocaListener2 VocaListener VocaListener2 VocaListener2 VocaListener VocaListener2 VocaListener2: A Singing Synthesis System Mimicking Voice Timbre Changes in Addition to Pitch and Dynamics
1
1 2 3 4 5 RESISTOR TUNABLE FILTER 6 LR-SERIES 1 1 2 3 4 5 6 7.1.1 1 1 1 RF1 CF1 RF2 CF2 INPUT 14 15 16 17 18 19 2 21 22 23 24 25 26 27 28 29 3 9 8 7 6 5 4 3 2 1 84 83 82 81 8 79 78 77 R R CF CF 56k R R
"CAS を利用した Single Sign On 環境の構築"
CAS 2 Single Sign On 1,3, 2,3, 2, 2,3 1 2 3 May 31, 2007 ITRC p. 1/29 Plan of Talk Brief survey of Single Sign On using CAS Brief survey of Authorization Environment using CAS 2 Summary May 31, 2007 ITRC
アナログ・デジタルの仕様とパフォーマンス特性の用語集
www.tij.co.jp Application Report JAJA127 Σ Σ 2 3 Σ 4 5 Σ Σ 2 2 1 1 Data Out 1 2 3 4 Data 1 2 3 4 Out Data Out 1 2 3 4 6 A CS B CLK 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 DOUT D 11 D 10 D 9 D 8 D 7 D 6 D 5
ETL Webinar
Logstash Filebeat ETL February 23rd, 2018 Kosho Owa, Principal Solution Architect Jun Ohtani, Developer Evangelist The Elastic Stack RESTFul Elasticsearch Elastic Stack Filebeat Beats The Beats family
"CAS を利用した Single Sign On 環境の構築"
CAS 2 SSO Authorization 1,3, 2,3, 2, 2,3 1 2 3 Central Authentication and Authorization Service (CAS 2 ) Web Application Single Sign On Authorization CAS 2 SSO/AuthZ Jan. 30 2007, p. 1/40 Plan of Talk
microSTATION Operation guide
J 2 - ii iii iv 1 1 2 4 7 8 9 10 11 1 5 6 3 2 10 15 9 11 12 13 14 3 7 6 5 4 3 2 1 4 5 PROGRAM OSC 1 MS1 (Multisample) Drum Kit MS2 (Multisample) MS3 (Multisample) Insert / Master / Total Effect IFX 1 MFX
ｽﾗｲﾄﾞ ﾀｲﾄﾙなし
(1) - E-Mail: [email protected] Y U V R G B (1/30 ) RGB / YUV = B G R V U Y 0.31 0.52 0.21 0.32 0.28 0.60 0.11 0.59 0.30 RGB YUV CCIR 601 4:4:4 4:2:2 4:2:0 Y Y Y U V U V U V YUVUV UV 4:2:0 4:2:2 (RGB8 )
467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 B =(1+R ) B +G τ C C G τ R B C = a R +a W W ρ W =(1+R ) B +(1+R +δ ) (1 ρ) L B L δ B = λ B + μ (W C λ B )
通信機用セラミックフィルタ（"セラフィル"）/セラミックディスクリミネータ
!! r Ceramic Filters (CERAFILr)/ Ceramic Discriminators for Communications Equipment Cat.No.P5-8 !! RoHS RoHS 2/65/EU RoHSWebRoHS http://www.murata.co.jp/info/rohs.html !! CONTENTS r 2 4 2 3 2 3 4 r 5