Konuşma aktivite detektörlerinde gürültülü dayanıklılığına etki eden faktörlerin incelenmesi

Bu makalede, literatürdeki bazı konuşma aktivite dedektörleri (VAD) değişen akustik gürültü koşullarına göre dayanıklılık performanslarını etkileyen faktörleri ölçmek amacıyla incelenmiş ve değişen gürültü koşullarına göre doğru tespit oranlarındaki değişimleri test edilerek ölçülmüştür. Bu kapsamda, VAD metotlarındaki karar aşamasında kullanılan eşik değerin sabit yada uyarlamalı olması, analiz penceresinin kısa yada uzun olması, birden fazla özellik vektörünün birlikte kullanımı gibi durumların sonuç performansa etkisi değerlendirilmiş ve karşılaştırmalı olarak analiz edilmiştir. Bu makalede incelenen dört farklı VAD dedektörünün üçü, karar sonucu üretirken kısa süreli analiz penceresi içerisindeki özellik vektörlerini kullanmakta iken, biri uzun vadeli spektral vektörlerin ölçüm sonucuna göre karar üretmektedir. Yine VAD detektörlerinin ikisi karar aşamasında sabit eşik kullanırken, diğer ikisi gürültüye göre uyarlamalı eşik kullanmaktadır. Analiz edilen VAD'lerin etkinliği, onları hem farklı akustik koşullar altında değerlendirmek ve hemde literatürde yer almış olan bir test verisi üzerinde test edebilmek için NOIZEUS corpus üzerinde test edilmiştir. Analiz edilen VAD'lerin testi sırasında, restoran, araba, sokak veya istasyon gibi [15-0dB] arasında çevresel arka plan gürültülerine sahip farklı türde giriş gürültülü konuşma sinyalleri test edilmiştir. Testler objektif test ölçüm metotları kullanılarak yapılmış ve herbir VAD metodunun tespit doğruluk oranı ölçülmüştür. Sonuçlar, herbir yöntemin, olumsuz çevresel koşullarda farklı dayanıklılık performansı verdiğini göstermiştir.

Anahtar Kelimeler:

konuşma aktivite algılama, konuşma analizi, arkaplan gürültüsü, uç nokta tespiti, sinyal-gürültü oranı

PDF

___

Javier Ramírez, Juan M Gorriz, José C. Segura, “Voice Activity Detection. Fundamentals and Speech Recognition System Robustness” DOI: 10.5772/4740, In book: Robust Speech Recognition and Understanding, Edited by Michael Grimm, I-Tech Education and Publishing, June 2007, ISBN: 978-3-902613-08-0
Javier Ramı́rez, José C Segura, Carmen Benı́tez, Ángel de la Torre, Antonio Rubio, “Efficient voice activity detection algorithms using long-term speech information”, Speech Communication, Volume 42, Issues 3–4,2004, Pages 271-287, ISSN 0167-6393,https://doi.org/10.1016/j.specom.2003.10.002.
L. R. Rabiner; M. R. Sambur, An algorithm for determining the endpoints of isolated utterances, The Bell System Technical Journal ( Volume: 54, Issue: 2, Feb. 1975 ), pp. 297 – 315, Feb. 1975, (ISSN: 0005-8580), DOI: 10.1002/j.1538-7305.1975.tb02840.x
A. Benyassine, E. Shlomot, Huan-Yu Su and E. Yuen, "A robust low complexity voice activity detection algorithm for speech communication systems," 1997 IEEE Workshop on Speech Coding for Telecommunications Proceedings, PA,USA, 1997, pp.97-98, doi: 10.1109/SCFT.1997.623914
K.Sakhnov, E.Verteletskaya and B. Simak (2009), Dynamical Energy-Based Speech/Silence Detector for Speech Enhancement Applications, Proceedings of the World Congress on Engineering 2009 Vol I, WCE 2009, July 1 - 3, London, U.K., ISBN: 978-988- 17012-5-1
R. G. Bachu, S. Kopparthi, B. Adapa and B. D. Barkana (2010), “Voiced/Unvoiced Decision for Speech Signals Based on ZeroCrossing Rate and Energy”, January, 2010, Advanced Techniques in Computing Sciences and Software Engineering, pp 279-282, 2010; DOI 10.1007/978-90-481-3660-5_47
Kyoung-Ho Woo, Tae-Young Yang, Kun-Jung Park, Chungyong Lee, “Robust voice activity detection algorithm for estimating noise spectrum”, February 2000, Electronics Letters 36(2):180 – 181, DOI: 10.1049/el:20000192
F. Beritelli, S. Casale, G. Ruggeri and S. Serrano, "Performance evaluation and comparison of G.729/AMR/fuzzy voice activity detectors," in IEEE Signal Processing Letters, vol. 9, no. 3, pp. 85-88,March2002, doi: 10.1109/97.995824
A. Benyassine, E. Shlomot, and H.-Y. Su, “ITU-T recommendation G.729 annex B: A silence compression scheme for use with G.729 optimized for V.70 digital simultaneous voice and data application,” IEEE Commun. Mag., vol. 35, pp. 64–73, Sept. 1997, https://www.itu.int/ITU-T/recommendations/rec.aspx?rec=11675 (for updated versions)
J. C. Junqua, B. Reaves, and B. Mark, “A study of endpoint detection algorithms in adverse conditions: Incidence on a DTW and HMM recognize,” in Proc. Eurospeech, 1991, pp. 1371-1374.
F. Lamel, R. Rabiner, E. Rosenberg, G. Wilpon: An improved endpoint detector for isolated word recognition, IEEE Trans. Acoust. Speech Signal Process., vol. 29, pp. 777-785, 1981.
B. Kotnik, Z. Kacic, B. Horvat: A multiconditional robust front-end feature extraction with a noise reduction procedure based on improved spectral subtraction algorithm, in Proc. 7th Europseech, pp. 197-200, 2001.
Tucker, R., “Voice activity detection using a periodicity measure”, IEE Proc. Inst. Elect. Eng, (Communications, Speech and Vision), Vol.139, No.4, pp. 377-380, August, 1992
J. Rouat, Y. C. Liu, and D. Morrisette, “A pitch determination and voiced/unvoiced decision algorithm for noisy speech,” Speech Commun., vol. 21, 1997. T. Nakatani, T. Irino, and P. Zolfaghari, “Dominance spectrum based V/UV classification and F estimation,” in Proc. Eurospeech ’03, 2003.
S. Ahmadi and A. S. Spanias, “Cepstrum-based pitch detection using a new statistical V/UV classification algorithm,” IEEE Trans. Speech Audio Process., vol. 7, no. 3, pp. 333–338, May 1999
Chengalvarayan, R. (1999). Robust energy normalization using speech/non-speech discriminator for German connected digit recognition, Proc. EUROSPEECH 1999, Budapest, Hungary, pp. 61–64.
BF Wu, KC Wang, Robust endpoint detection algorithm based on the adaptive band partitioning spectral entropy in adverse environments. IEEE Transactions Speech Audio Processing 13, 762–775 (2005)
S. O. Sadjadi and J. H. L. Hansen, "Unsupervised Speech Activity Detection Using Voicing Measures and Perceptual Spectral Flux," in IEEE Signal Processing Letters, vol. 20, no. 3, pp. 197-200, March 2013. doi: 10.1109/LSP.2013.2237903
Nemer, E.; Goubran, R.; Mahmoud, S. (2001). Robust voice activity detection using higherorder statistics in the lpc residual domain, IEEE Trans. Speech Audio Processing, vol. 9, no. 3, pp. 217–231.
B. Kingsbury, G. Saon, L. Mangu, M. Padmanabhan, R. Sarikaya: Robust speech recognition in noisy environments: the 2001 IBM SPINE evaluation system, Proc. ICASSP, pp. 53-56, 2002.
T. Kristjansson, S. Deligne, P. Olsen: Voicing features for robust speech detection, Proc. Interspeech, pp. 369-372, 2005.
Marzinzik, M.; Kollmeier, B. (2002). Speech pause detection for noise spectrum estimation by tracking power envelope dynamics, IEEE Trans. Speech Audio Processing, vol. 10, no. 6, pp. 341–351.
S. A. McClellan and J. D. Gibson, “Spectral entropy: An alternative indicator for rate allocation, ” in IEEE Int. Conf. on Acoustics, Speech, Signal Processing, (Adelaide, Australia), pp. 201-204, Apr. 1994.
T Kristjansson, S Deligne, P Olsen, Voicing features for robust speech detection (INTERSPEECH, 2005), pp. 369–372
J. A. Haigh and J. S. Mason, “Robust voice activity detection using cepstral features,” in Proc. IEEE TENCON,'93. Proceedings. Computer, Communication, Control and Power Engineering.1993 IEEE Region 10 Conference on, 1993, China, pp. 321-324 vol.3.
A. M. Noll, “Cepstrum pitch determination,” J. Acoust. Soc. Amer., vol. 41, 293-309, Feb. 1967.
J. A. Haigh, J. S. Mason, "Robust voice activity detection using cep¬stral features", Proc. of TENCON '93. IEEE Region 10 Interna¬tional Conference on Computers, Communications and Automa¬tion, Beijing, China, 1993.
K. Chung, S. Y. Oh, "Voice Activity Detection Using an Improved Unvoiced Feature Normalization Process in Noisy Environments", Wireless Personal Communications, vol. 89, no. 3, pp. 1-13, 2015.
S Ahmadi, AS Spanias, Cepstrum-based pitch detection using a new statistical V/UV classification algorithm. IEEE Transactions on Speech Audio Processing 7, 333–338 (1999)
J. Stegmann and G. Schroder, "Robust voice-activity detection based on the wavelet transform," in Speech Coding for Telecommunications Proceeding, 1997, 1997 IEEE Workshop on, 1997, pp. 99-100.
S. H. Chen, H. Te Wu, Y. Chang, T. K. Truong, "Robust voice activity detection using perceptual wavelet-packet transform and Teager energy operator", Pattern Recognition Letters, vol. 28, no. 11, pp. 1327-1332, 2007.
Jongseo Sohn, Nam Soo Kim and Wonyong Sung, "A statistical model-based voice activity detection," in IEEE Signal Processing Letters, vol. 6, no. 1, pp. 1-3, Jan. 1999, doi: 10.1109/97.736233
J. Sohn, N. S. Kim, and W. Sung, “A statistical model-based voice activity detection,” IEEE Signal Process. Lett., vol. 6, pp. 1–3, Jan. 1999.
Y. D. Cho, K. Al-Naimi, and A. Kondoz, “Improved statistical voice activity detection based on a smoothed statistical likelihood ratio,” in Proc. IEEE ICASSP’01, vol. 2, Salt Lake City, UT, 2001, pp. 737–740.
R. Tahmasbi and S. Rezaei, “Change point detection in GARCH models for voice activity detection,” IEEE Trans. Audio, Speech, Lang. Process., vol. 16, no. 5, pp. 1038–1046, Jul. 2008.
A. Davis, S. Nordholm, and R. Togneri, “Statistical voice activity detection using low-variance spectrum estimation and an adaptive threshold,” IEEE Trans. Audio, Speech, Lang. Process., vol. 14, no. 2, pp. 412–424, Mar. 2006.
J. Chang, N. K. Kim, and S. K. Mitra, “Voice activity detection based on multiple statistical models,” IEEE Trans. Signal Process., vol. 54, no.6, pp. 1965-1976, Jun. 2006.
J. Haigh, J. Mason: A voice activity detector based on cepstral analysis, Proc. Eurospeech, pp. 1103-1106, 2003.
D. Ying, Y. Yan, J. Dang, F. Soong: Voice Activity Detection Based on an Unsupervised Learning Framework, IEEE Trans. Audio Speech and Lang. Process., vol. 19, pp. 2624-2633, 2011.
F. Eyben, F. Weninger, S. Squartini, B. Schuller: Real-life voice activity detection with LSTM Recurrent Neural Networks and an application to Hollywood movies, in Proc. ICASSP, pp.483-487, 2013.
G. Ferroni, R. Bonfigli, E. Principi, S. Squartini, F. Piazza, "A Deep Neural Network approach for Voice Activity Detection in multi-room domestic scenarios", International Joint Conference on Neu¬ral Networks (IJCNN), Killarney, Ireland, 2015.
F. Bie, Z. Zhang, D. Wang, T. F. Zheng, "DNN-based Voice Activity Detection for Speaker Recognition", CLST Technical Report, pp. 1-11, 2015.
Z. Ali, M. Talha, "Innovative Method for Unsupervised Voice Activ¬ity Detection and Classification of Audio Segments", IEEE Access, vol. 6, pp. 15494-15504, 2018.
GSM 06.94. (1999, Feb.) Digital cellular telecommunication system (Phase 2+); voice activity detector VAD for adaptive multi rate (AMR) speech traffic channels; general description. ETSI, Tech. Rep. V.7.0.0. [Online]. Available: http://www.etsi.org.
GSM 06.94. (1999, Feb.) Digital cellular telecommunication system (Phase 2+); voice activity detector VAD for adaptive multi rate (AMR) speech traffic channels; general description. ETSI, Tech. Rep. V.7.0.0,1999
Yiu-Kei Lau and Chok-Ki Chan, "Speech recognition based on zero crossing rate and energy," in IEEE Transactions on Acoustics, Speech, and Signal Processing, vol. 33, no. 1, pp. 320-323, February 1985. doi: 10.1109/TASSP.1985.1164503
Charles W. Brokish, Michele Lewis, A-Law and µ-Law Companding Implementations Using the TMS320C54x, Texas ınstruments, 1997
Javier Ramı́rez, José C Segura, Carmen Benı́tez, Ángel de la Torre, Antonio Rubio, “Efficient voice activity detection algorithms using long-term speech information”, Speech Communication, ISSN 0167-6393,Volume 42, Issues 3–4,2004,pp: 271-287, https://doi.org/10.1016/j.specom.2003.10.002
A. Benyassine, E. Shlomot, and H.-Y. Su, “ITU-T recommendation G.729 annex B: A silence compression scheme for use with G.729 optimized for V.70 digital simultaneous voice and data application,” IEEE Commun. Mag., vol. 35, pp. 64–73, Sept. 1997
F. Beritelli, S. Casale, and A. Cavallaro, “A robust voice activity detector for wireless communications using soft computing,” IEEE J. Select Areas Commun., vol. 16, pp. 1818–1829, Dec. 1998.
Wu, J., Zhang, X. An efficient voice activity detection algorithm by combining statistical model and energy detection. EURASIP J. Adv. Signal Process. 2011, 18 (2011). https://doi.org/10.1186/1687-6180-2011-18
Meduri, S., Ananth, R., “A Survey and Evaluation of Voice Activity Detection Algorithms”, MsC Thesis, Karlskrona, Sweden, June, 2011
Loizou, P, “NOIZEUS: A noisy speech corpus for evaluation of speech enhancement algorithms”, Speech Communication, 49, 588-601, (2017).
. Hirsch, HG, Pearce, D, “The Aurora experimental framework for the performance evaluation of speech recognition systems under noisy conditions”, ISCA Tutorial and Research Workshop (ITRW) ASR2000, September 2000