Reading Groups
Machine Learning
Tone
Machine Learning
November 15, 2004: Language Models
November 22, 2004: Deviation from Randomness
- Main Paper
- More theoretical
- Random indexing
- Maxent
- Related papers by similar authors can be found at the same site
November 29, 2004: Iterative Residual Rescaling
- Iterative Residual Rescaling: An Analysis and Generalization of {LSI}
December 6, 2004: Spoken Document Retrieval
- Document Expansion for Speech Retrieval. Amit Singhal, Fernando Pereira. ACM SIGIR'99, pages 26-33, 1999.
- SCAN: Designing and Evaluating User Interfaces to Support Retrieval from Speech Archives. Steve Whittaker, Julia Hirschberg, John Choi, Don Hindle, Fernando Pereira, Amit Singhal. ACM SIGIR'99, 34-41, 1999.
- Search conversational speech
December 13, 2004: Drago Radev's visit
- Catching the Drift: Probabilistic Content Models, with Applications to Generation and Summarization, Barzilay and Lee (HLT-NAACL 2004).
- E. Hovy and Chen-Yew Lin, Automated Text Summarization in SUMMARIST, in Advances in Automatic Text Summarization, 1999.
- Dragomir R. Radev; Eduard Hovy; Kathleen McKeown, Introduction to the Special Issue on Summarization
- Automatic Summarization of Open-Domain Multiparty Dialogues in Diverse Genres
January 3, 2005: NIPS report out
January 10, 2005: Dialogue Act Tagging
- A. Venkataraman, L. Ferrer, A. Stolcke, & E. Shriberg (2003), Training a Prosody-based Dialog Act Tagger from Unlabeled Data. Proc. IEEE Intl. Conf. on Acoustics, Speech and Signal Processing, Hong Kong.
- Shriberg, E., Bates, R., Stolcke, A., Taylor, P., Jurafsky, D., Ries, K., Coccaro, N., Martin, R., Meteer, M., Van Ess-Dykema, C. (1998). Can Prosody Aid the Automatic Classification of Dialog Acts in Conversational Speech? In M. Swerts and J. Hirschberg (eds.) Special Double Issue on Prosody and Conversation. Language and Speech 41(3-4), 439-487.
- Dialog Act Classification With The Help Of Prosody (1996)
Int. Conf. on Spoken Language Processing, volume 3, pages 1728--1731, Philadelphia, 1996. Mast et al
- Helen Wright, Massimo Poesio, and Stephen Isard. 1999. Using high level dialogue information for dialogue act recognition using prosodic features. In Proceedings of the ESCA Workshop on Prosody and Dialogue, Eindhoven, September 1999.
January 17, 2005: Latent Dirichlet Allocation: Techniques and Application
- Latent Dirichlet Allocation (2003), David M. Blei, Andrew Y. Ng, Michael I. Jordan, In: NIPS*14. (2002)
- On an Equivalence between PLSI and LDA (2003), Mark Girolami, Ata Kaban, In Proceedings of SIGIR 2003.
- Integrating Topics and Syntax, Thomas L. Griffiths, Mark Steyvers, David M. Blei, Joshua B. Tenenbaum, In: Advances in Neural Information Processing Systems, 17
- D. Blei, Michael, and M. I. Jordan. Modeling annotated data. Proceedings of the 26th annual intetational ACM SIGIR conference
Monday February 7, 2005: Recognizing and Recognition with Tone and Intonation
- Chen & H-Johnson 2004, How Prosody improves Speech Recognition 2004
- Xuejing Sun 2002, Pitch Accent Prediction Using Ensemble Machine Learning
- Kurt E. Dusterhoff. Automatic intonation analysis using acoustic data. In Proceedings, ESCA TRW on Dialogue and Prosody, Eindhoven, 1999
February 14, 2005: Max-Margin Methods
- Max-Margin Parsing, Ben Taskar, Dan Klein, Mike Collins, Daphne Koller and Christopher Manning, Empirical Methods in Natural Language Processing - EMNLP '2004, p. 1-8
- Hidden Markov Support Vector Machines
- Large Margin Methods for Label Sequence Learning
February 21, 2005: Discourse Structure
- Derrick Higgins, Jill Burstein, Daniel Marcu, and Claudia Gentile (2004).
Evaluating Multiple Aspects of Coherence in Student Essays In Proceedings
of the Human Language Technology and North American Association for
Computational Linguistics Conference (HLT/NAACL 2004), May 2-5, Boston,
MA.
- Daniel Marcu and Abdessamad Echihabi (2002). An Unsupervised Approach to Recognizing Discourse Relations. Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics (ACL-2002), Philadelphia, PA, July 7-12.
- Radu Soricut and Daniel Marcu (2003). Sentence Level Discourse Parsing using Syntactic and Lexical Information. Proceedings of the Human Language Technology and North American Association for Computational Linguistics Conference (HLT/NAACL), May 27-June 1, Edmonton, Canada.
- Jill Burstein, Daniel Marcu, and Kevin Knight (2003). Finding the WRITE Stuff: Automatic Identification of Discourse Structure in Student Essays. IEEE Intelligent Systems, pp. 32-39, Jan/Feb, 2003
Tone
November 13, 2004: PENTA Model
November 30, 2004: Tone & Focus
- http://home.uchicago.edu/~xuyi/publications.html
- Xu, Y. (1999). Effects of tone and focus on the formation and alignment of F0 contours. Journal of Phonetics, 27: 55-105.
- Optional
- Xu, Y. (1997). Contextual tonal variations in Mandarin. Journal of Phonetics, 25: 61-83.
December 2, 2004
- Linguistics tone talk
Thursday, December 2
4 p.m. Stuart 105
Contrast enhancement in narrow focus and clear speech
Rajka Smiljanic
Northwestern University
ABSTRACT
In this talk, I will present data from two projects that examined how
the phonological properties of a language (e.g., presence vs. absence
of a phonemic contrast, the number of contrastive segments, phonemic
vs. allophonic contrasts, etc.) and pragmatic factors (e.g.,
hyperarticulated/clear speech, narrow focus) condition speech
production. I will first discuss the interaction between pragmatic
narrow/contrastive focus and lexical contrast as factors that
condition variation in pitch and duration in Croatian and Serbian.
These two closely related dialects differ in the presence or absence
of lexical pitch-accent and vowel length contrasts. It is found that
lexical pitch-accent and vowel length contrasts influence the
expression of pragmatic focus: the phonemic contrasts are enhanced in
narrow focus rather than uniformly made more prominent.
December 9, 2004: Tone & Focus
- Yi Xu's publications
- Xu, Y. (1999). Effects of tone and focus on the formation and alignment of F0 contours. Journal of Phonetics, 27: 55-105.
- Xu, Y. (1997). Contextual tonal variations in Mandarin. Journal of Phonetics, 25: 61-83.
January 6, 2005
- Chapters 1-3, Gussenhoven, phonology of tone & intonation
January 13, 2005
January 31, 2005: STEM-ML
- Quantitative measurement and prediction of prosodic strength in Mandarin. Greg
Kochanski, Chilin Shih and Hongyan Jing (2003). Speech Communication. V. 41, No. 4, pp. 625-645.
- Hierarchical Structure and Word Strength Prediction of Mandarin Prosody. Greg Kochanski, Chilin Shih and Hongyan Jing (2003). International Journal of Speech Technology, 6 (1), pp. 33-43.
- Automated Modelling of Chinese Intonation in Continuous Speech. Greg Kochanski and Chilin Shih (2001). Eurospeech, pp. 911-914, Aalborg, Denmark
February 10, 2005
February 17, 2005: Pitch Accent Prediction
- I. Bulyko and M. Ostendorf. "A Bootstrapping Approach to Automating Prosodic Annotation for Constrained Domain Synthesis", in Proceedings of the IEEE Workshop on Speech Synthesis, 2002.
- K. Ross and M. Ostendorf, ``Prediction of Abstract Prosodic Labels for Speech Synthesis,'' Computer, Speech and Language, Vol. 10, No. 3, July 1996, pp. 155-185.
- Shimei Pan and Julia Hirschberg (2000), Modeling Local Context for Pitch Accent Prediction (PostScript file) or (PDF file) , ACL'2000, Hong Kong.
February 24, 2005: Tone Recognition for Chinese
- Y. Qian, Tan Lee and Frank K. Soong "Tone Information as a Confidence Measure for Improving Cantonese LVCSR",8th International Conference on Spoken Language Processing, vol. III,pp.1965 - 1968, Jeji Island, Korea, October 2004.
- Yao Qian, Tan Lee and Frank K. Soong, "Use of tone information in continuous Cantonese speech recognition", in Proceedings of Speech Prosody 2004, pp.587- 590, Nara, Japan, March 2004
- C. Wang and S. Seneff, "Improved Tone Recognition by Normalizing For Coarticulation and Intonation Effects." Proc. 6th International Conference on Spoken Language Processing, Beijing, China October 2000. (PDF)
- Tone Articulation Modeling for Mandarin Spontaneous Speech Recognition
Jian-lai Zhou; Ye Tian; Yu Shi; Chao Huang; Eric Chang
May 2004
March 10, 2005 and March 17: Implementing Pitch Target Approximation
- Xuejing Sun's thesis: pp. 90-190
June 14, 2006: Multi-stream DBNs for Tone Recognition
- Jeff Bilmes' paper on recognizing tonal phones with DBNs
- Xin Lei, Gang Ji, Tim Ng, Jeff Bilmes, and Mari Ostendorf DBN-Based Multi-stream Models for Mandarin Toneme Recognition (gzipped ps or pdf) IEEE Int. Conf. on Acoustics, Speech, and Signal Processing, Philadelphia, PA, March 2005
June 21, 2006: Minimal Supervision for Pitch Accent
- Wong et al's paper on ME, self and co-training
- Using Weakly Supervised Learning to Improve Prosody Labeling,
D. Wong, M. Ostendorf and J. Kahn