speech separation for recognition and enhancementdpwe/talks/darpa-2011-10-27.pdfspeech separation -...
TRANSCRIPT
Speech Separation - Dan Ellis 2011-10-27 /181
1. Speech in the Wild2. Separation by Space3. Separation by Pitch4. Separation by Model
Speech Separation forRecognition and Enhancement
Dan EllisLaboratory for Recognition and Organization of Speech and Audio
Dept. Electrical Eng., Columbia University, NYand
International Computer Science Institute, Berkeley CA
[email protected] http://labrosa.ee.columbia.edu/
Speech Separation - Dan Ellis 2011-10-27 /18
1. Speech in the Wild
2
• The world is cluttered sound is transparentmixtures are inevitable
• Useful information is structured by ‘sources’specific definition of a ‘source’:intentional independence
Speech Separation - Dan Ellis 2011-10-27 /18
Speech in the Wild: Examples
• Multi-party discussions
• Ambient recordings
• Applications:communications o robots o lifelogging/archives
3
Speech Separation - Dan Ellis 2011-10-27 /18
Recognizing Speech in the Wild• Current ASR relies on low-D representations
e.g. 13 dimensional MFCC features every 10ms
• We need separation!
4
time / s
level / dB
ICSI Meeting Room excerpt
freq
/ kHz
MFCC!based resynthesis
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
freq
/ kHz
0
1
2
3
4
!60
!40
!20
0very successful for clean speech!inadequate for mixtures
Speech Separation - Dan Ellis 2011-10-27 /18
2. Speech Separation• How can we separate speech information?
5
Select / Enhance
Analyze
Application
Noisy Speech
CleanedSpeech
(features)• Recognition• Listening• ...
• Spatial• Pitch• Speech probs• ...
• T-F masking• Weiner filtering• Reconstruction
Speech Separation - Dan Ellis 2011-10-27 /18
Separation by Spatial Info• Given multiple microphones,
sound carries spatial information about source• E.g. model interaural spectrum of each source
as stationary level and time differences:
• e.g. at 75°, in reverb:
6
IPD IPD residual ILD
IPD ILDIPD residual
Speech Separation - Dan Ellis 2011-10-27 /18
Model-Based EM Source Separation and Localization (MESSL)
can model more sources than sensors7
Mandel et al. ’10
Assign spectrogram pointsto sources
Re-estimatesource parameters
Speech Separation - Dan Ellis 2011-10-27 /18
MESSL Results• Modeling uncertainty improves results
tradeoff between constraints & noisiness
8
2.45 dB2.45 dB
9.12 dB9.12 dB
0.22 dB0.22 dB
-2.72 dB-2.72 dB
12.35 dB12.35 dB8.77 dB8.77 dB
Ground truth masks
!40 !20 0 20 400
20
40
60
80
100
Target!to!masker ratio (dB)
Perce
nt co
rrect
HumanDP!OracleOracleOracleAllRevMixes
Algorithmic masks
!40 !20 0 20 400
20
40
60
80
100
Target!to!masker ratio (dB)
HumanSawadaMoubaMESSL!GMESSL!!!DUETMixes
• Helps with recognitiondigits accuracy
Speech Separation - Dan Ellis 2011-10-27 /18
3. Separation by Pitch• Voiced syllables have near-periodic “pitch”
perceptually salient
lost in MFCCs
• Can we track pitch & use it for separation?... and other speech tasks?
9
!"#$%&$'()#*"&+$,-"#$'(!$./-#01232'(4-**2&2#52.
1232'(6-**2&2#52(5$#(72'8(-#(.82257(.20&20$,-"#
1-.,2#(*"&(,72(9%-2,2&(,$'/2&:
Brungart et al.’01
Speech Separation - Dan Ellis 2011-10-27 /18
Noise-Robust Pitch Tracking• Important for voice detection & separation• Based on channel selection Wu, Wang & Brown ’03
pitch from summary autocorrelation over “good” bands
10
trained classifier decides which channels to include
BS Lee & Ellis ’12
Speech Separation - Dan Ellis 2011-10-27 /18
Noise-Robust Pitch Tracking• Channel-based classifiers
learn domain channel/noise characteristicsthen separate, or derive features for recognition
• Only works for pitched soundsneed a broader description of the speech source...
11
Freq
/ kH
z
0
2
4
Cha
nnel
s
2040
0
12.5
25
0
12.5
25
perio
d / m
s
−40−20020
08_rbf_pinknoise5dB
0 1 2 3 4 time / sec
CS-GT
Speech Separation - Dan Ellis 2011-10-27 /18
4. Separation by Models• If ASR is finding best-fit parameters
argmax P(W | X) ...• Recognize mixtures with Factorial HMM
model + state sequence for each voice/sourceexploit sequence constraints, speaker differences
separation relies on detailed speaker model12
Varga & Moore, ’90Hershey et al., ’10
model 1
model 2
observations / time
Speech Separation - Dan Ellis 2011-10-27 /18
Eigenvoices• Idea: Find
speaker model parameter space
generalize without losing detail?
• Eigenvoice model:
89,600 dimensional space13
Kuhn et al. ’98, ’00Weiss & Ellis ’10
Speaker modelsSpeaker subspace bases
µ = µ̄ + U w + B hadapted mean eigenvoice weights channel channelmodel voice bases bases weights
Freq
uenc
y (kH
z)
Mean Voice
b d g p t k jh ch s z f th v dh m n l r w y iy ih eh ey ae aaaw ay ah aoowuw ax
2
4
6
8
Freq
uenc
y (kH
z)
Eigenvoice dimension 1
b d g p t k jh ch s z f th v dh m n l r w y iy ih eh ey ae aaaw ay ah aoowuw ax
2
4
6
8
Freq
uenc
y (kH
z)
Eigenvoice dimension 2
b d g p t k jh ch s z f th v dh m n l r w y iy ih eh ey ae aaaw ay ah aoowuw ax
2
4
6
8
Freq
uenc
y (kH
z)Eigenvoice dimension 3
b d g p t k jh ch s z f th v dh m n l r w y iy ih eh ey ae aaaw ay ah aoowuw ax
2
4
6
8
50
40
30
20
10
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
Speech Separation - Dan Ellis 2011-10-27 /18
Eigenvoice Speech Separation
14
Speech Separation - Dan Ellis 2011-10-27 /18
Eigenvoice Speech Separation• Eigenvoices for Speech Separation task
speaker adapted (SA) performs midway between speaker-dependent (SD) & speaker-indep (SI)
15
SI
SA
SD
Mix
Speech Separation - Dan Ellis 2011-10-27 /18
Spatial + Model Separation• MESSL + Eigenvoice “priors”
16
Weiss, Mandel & Ellis ’11
Speech Separation - Dan Ellis 2011-10-27 /18
Summary• Speech in the Wild
... real, challenging problem
... applications in communications, lifelogs ...
• Speech Separation... by generic properties (location, pitch)... via statistical models
• Recognition and Enhancement... separate-then-X, or integrated solution?
17
Speech Separation - Dan Ellis 2011-10-27 /18
References• John Hershey, Steve Rennie, Pedr Olsen, Trausti Kristjansson, “Super-human multi-talker speech
recognition: A graphical modeling approach,” Computer Speech & Lang. 24 (1), 45-66, 2010.
• Jon Barker, Martin Cooke, Dan Ellis, “Decoding Speech in the Presence of Other Sources,” Speech Communication 45(1): 5-25, 2005.
• R. Kuhn, J. Junqua, P. Nguyen, N. Niedzielski, “Rapid speaker adaptation in eigenvoice space,” . IEEE Tr. Speech & Audio Proc. 8(6): 695–707, Nov 2000.
• Byung-Suk Lee & Dan Ellis, “Noise-robust pitch tracking by trained channel selection,” submitted to ICASSP, 2012.
• Michael Mandel, Ron Weiss, Dan Ellis, “Model-Based Expectation-Maximization Source Separation and Localization,” IEEE Tr. Audio, Speech, Lang. Proc. 18(2): 382-394, Feb 2010.
• A. Varga and R. Moore, “Hidden markov model decomposition of speech and noise,” ICASSP-90, 845–848, 1990.
• Ron Weiss & Dan Ellis, “Speech separation using speaker-adapted Eigenvoice speech models,” Computer Speech & Lang. 24(1): 16-29, 2010.
• Ron Weiss, Michael Mandel, Dan Ellis, “Combining localization cues and source model constraints for binaural source separation,” Speech Communication 53(5): 606-621, May 2011.
• Mingyang Wu, DeLiang Wang, Guy Brown, “A multipitch tracking algorithm for noisy speech,” IEEE Tr. Speech & Audio Proc. 11(3): 229–241, May 2003.
18