- Geographical location
- Workshop venue (facilities, meeting rooms, network access during the workshop, audio/visual equipment)
- Accommodation -- hotel availability and pricing
- Meals
- Transportation options -- major airports, logistics, visas
- Climate
- Previous workshops have been held in the month of December.
- Please be aware of any other related conferences in the same time frame.
- Approximately how much will participants need to pay to attend, including accommodation and meals as well as registration?
- Estimated budget for e.g. 100/150 participants and expected sponsorships. This should include venue costs, administration, banquet, coffee breaks, publication costs, etc. (Note, unlike larger conferences such as ICASSP, Interspeech, smaller workshops do not have sufficient registration fees to cover the total workshop costs. The IEEE Signal Processing Society is particularly motivated to see that both workshop and conference costs are held to levels which allow the diverse membership equal opportunities to participate without registration costs being a major barrier).
- Typically 15-20 people with a mix of academic and industrial researchers
- Program committees are encouraged to reach a good balance of SLTC and non-SLTC members, for example approximately a 50/50 split
- Who will be in charge of organizing the workshop, and how will finances be handled (e.g., will participants be able to pay by credit card)?
- Paper submission, notification of acceptance, proposals for demonstrations, early registration
- Reception, talks, posters, demo session, banquet, etc.
If you would like to be the organizer(s) of ASRU 2015, please send the Workshop Sub-Committee a draft proposal before April 25, 2014. (Point of contact: Nick Campbell or George Saon). Proposals will be evaluated by the SPS SLTC, with a decision expected in June.
The organizers of the ASRU workshop do not have to be SLTC members, and we encourage submissions from all potential organizers. So we encourage you to distribute this call for proposals far and wide to invite members of the speech and language community at large to submit a proposal to organize the next ASRU workshop.
For more information on the most recent workshops, please see the following:
- For information about ASRU 2013 in Olomouc, Czech Republic
- For information about ASRU 2011 on the Big Island of Hawaii
- For information about ASRU 2009 in Merano, Italy
Speech Synthesis Perfects Everyone’s Singing
Minghui Dong, Nancy Chen, Haizhou Li
SLTC Newsletter, February 2014
Singing is more expressive than speaking. While singing is popular, singing well is nontrivial. This is especially true for songs that require high vocal skills. A singer needs to overcome two challenges among others - to sing in the right tune and at the correct rhythm. Even professional singers need intensive practice to perfect their vocal skills and to proficiently present particular singing styles, such as vibrato and resonance tuning. Recently, the Institute for Infocomm Research (I2R) in Singapore has developed a technology called Speech2Singing, which converts the singing voice of non-professional singers (or even spoken utterances) into perfect singing.
The human voice includes three essential elements: content, prosody and timbre. Content is concerned with the literal meaning of language conveyed by the voice. Prosody consists of pitch, duration (timing) and loudness of voice. Prosody characterizes the emotion and expressiveness of one’s voice. For the case of singing voice, prosody is often referred to as melody (a combination of pitch and rhythm). Timbre, on the other hand, determines the identity of a person’s voice. I2R’s Speech2Singing technology keeps the content and timbre of the voice unchanged, but modifies the prosody of the voice into the correct melody to perfect the singing voice of the user. I2R’s Speech2Singing works as follows: Singing voices of professional singers are recorded as model voice templates and stored in the database. When a user subsequently sings a song or reads the lyrics, the recorded voice of each line is compared with the corresponding line stored in the database. The user’s vocal signal is first decomposed into feature (including pitch) sequences. An enhanced singing voice is then synthesized from the adjusted feature sequence, which contains the correct pitch and timing information. To obtain the correct timing for the user’s voice, speech recognition technology is used to identify the phonetic units from both the model voice and the user’s voice. The timing information of the user’s voice is adjusted to match that of the model voice by means of dynamic time warping [1]. The correct pitch information is directly derived from the model singing voice. Finally, the reconstructed time-synchronous singing voice is overlaid with background music.
Two existing technologies related to Speech2Singing are Auto-Tune [2] and score-based conversion [3]. The popular method of Auto-Tune alters the user’s pitch to the closest note in a pre-defined scale. While it generally works well for singing voice where the melody is not far off-tune, it is not suitable for converting spoken voice. In contrast to Auto-tune, I2R's method uses a reference melody to guide the conversion, so that corrections can still be made even when the melody is completely off-tune.
Although the score-based conversion method in [3] uses musical scores as reference for the change of melody (pitch and timing), the reference melody is generated with mathematical models, making the synthesized singing voice less natural than when using human melody as a reference. I2R’s approach uses a professional singer’s melody as a reference model so that every single detail of the melody, such as the pitch envelope and vibrato (a regular, pulsating change in pitch), is perfectly preserved and imposed onto the synthesized singing voice. In addition, since the user’s voice can be mapped to the professional singer’s voice, I2R’s method allows the timing change of each syllable of the user’s voice to match that of the professional singers’ voice.
I2R has implemented this Speech2Singing technology in mobile devices such as smart-phones and tablets. This is the first software that automatically changes a user’s speech into natural singing voice. The technology has been showcased in various occasions such as I2R’s annual TechFest [4] and A*STAR’s MediaExploit [5]; it has also drawn attention from local and international media, such as AFP [6], C-Net [7], MediaCorp [8]. ‘Sing for Singapore’ was the first release of Speech2Singing to the public during 2013 Singapore’s National Day (Figure 1) [9,10] with iOS version in AppStore [11] and Android version in Google Play [12].
In 1961, an IBM 7094 became the first computer to sing (the song was “Daisy Bell”). Ever since, singing synthesis technology has progressed tremendously. Similar to Photoshop that perfects graphics, Speech2Singing technology helps perfect singing vocals.
Figure 1: Screen shots of the NDP 2013 App.
References:
[1] L Cen, M Dong, P Chan, Template-based Personalized Singing Voice Synthesis, IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2012.
[2] http://www.antarestech.com/
[3] T. Saitou, M. Goto, M. Unoki and M. Akagi, "Speech-to-Singing Synthesis: Vocal conversion from speaking voices to singing voices by controlling acoustic features unique to singing voices," Proc. Proc. 2007 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA2007), pp. 215–218, 2007.
[4] http://asia.cnet.com/scientists-showcase-innovations-at-singapores-techfest-62218326.htm
[7] http://asia.cnet.com/videos/iframe-embed/46728161/
[8] http://www.etpl.sg/innovation-offerings/technologies-for-license/tech-offers/1908
[9] https://www.facebook.com/i2r.research/posts/679594635401672
[10] http://www.etpl.sg/find-us-at/news/in-the-news/article/119
[11] https://itunes.apple.com/sg/app/ndp-2013-mobile-app/id524388683?mt=8
[12] https://play.google.com/store/apps/details?id=com.massiveinfinity.slidingmenu&hl=en
Minghui Dong is a Scientist in the Department of Human Language Technology at Institute for Infocomm Research, Singapore. His research interests include speech synthesis, singing voice synthesis, and voice conversion.
Nancy Chen is a Scientist in the Department of Human Language Technology at Institute for Infocomm Research, Singapore. Her research interests include keyword search, pronunciation modeling, speech summarization, and computer-assisted language learning. For more information: http://alum.mit.edu/www/nancychen.
Haizhou Li is the Head of the Department of Human Language Technology at Institute for Infocomm Research, Singapore. He is also a Conjoint Professor at the University of New South Wales, Australia.
1 Billion Word Language Modeling Benchmark
Ciprian Chelba
SLTC Newsletter, February 2014
We just released a LM benchmark at: https://code.google.com/p/1-billion-word-language-modeling-benchmark/ and would like to advertise it to the speech community.
The purpose of the project is to make available a standard training and test setup for language modeling experiments.
The training/held-out data was produced from a download at statmt.org using a combination of Bash shell and Perl scripts distributed here.
This also means that your results on this data set are reproducible by the research community at large.
Besides the scripts needed to rebuild the training/held-out data, it also makes available log-probability values for each word in each of ten held-out data sets, for each of the following baseline models:
- unpruned Katz (1.1B n-grams),
- pruned Katz (~15M n-grams),
- unpruned Interpolated Kneser-Ney (1.1B n-grams),
- pruned Interpolated Kneser-Ney (~15M n-grams)
ArXiv paper: http://arxiv.org/abs/1312.3005
Happy benchmarking!
An Overview of ASRU 2013
Tara N. Sainath and Jan (Honza) Cernocky
SLTC Newsletter, February 2014
The Automatic Speech Recognition and Understanding Workshop (ASRU) was recently hosted in Olomouc, Czech Republic from December 8-12, 2013. Each day of the workshop focused on a specific current theme that has become popular amongst ASR researchers. Below, we highlight the 4 days of the workshop, and touch on some interesting papers, in more detail.
Day 1 – Neural Network
In the past few years, deep learning has become the de-facto approach for acoustic modeling in ASR, showing tremendous improvements between 10-30% relative over alternative acoustic modeling approaches across a variety of LVCSR tasks [1].
The stage was set by a keynote on the Physiological Basis of Speech Processing: From Rodents to Humans by Christoph Schreiner (UC San Francisco), followed by invited talks on Multilayer perceptrons for speech recognition: There and Back Again by Nelson Morgan (International Computer Science Institute), and Context-Dependent Deep Neural Networks for Large Vocabulary Speech Recognition: From Discovery to Practical Systems by Frank Seide (Microsoft Research Asia).
There were numerous interesting papers in this section as well. For example [2] explored using speaker-indentity vectors with DNNs, obtaining impressive results on Switchboard. In addition, [3] introduced a semi-supervised training strategy for DNNs. On the surprise language task (Vietnamese) from the BABEL project, they were able to obtain a 2.2% absolute improvement in WER compared to a system built only on fully transcribed data. Finally, [4] looked at moving benefits of DNNs back into GMM modeling techniques, including having deep (multiple layers) and wide (multiple) models and by sharing model parameters. Making these changes, the authors find that the performance of GMMs comes closer to DNNs on TIMIT.
Day 2 – Limited Resource
The second day focused on limited resources for ASR.
Invited talks were given on Building Speech Recognition Systems with Low Resources (by Tanja Schultz, Karlsruhe Institute of Technology) and Unsupervised Acoustic Model Training with Limited Linguistic Resources (by Lori Lamel, CNRS-LIMSI). Mary Harper (IARPA) delivered a keynote on The Babel Program and Low Resource Speech Technology, which was followed by invited talks on Zero to One Hour of Resources: A Self-organizing Unit Approach to Training Speech Recognizers (by Herb Gish, Raytheon BBN Technologies), and Recent Progress in Unsupervised Speech Processing, by Jim Glass (Massachusetts Institute of Technology).
The work in [5] looked at automatically learning a pronunciation lexicon, by starting with a small seed lexicon and then learning pronunciations of words by transcribing speech at the word level. Experiments on a Switchboard task show that the proposed lexicon learning method achieves a WER similar to using a fully handcrafted lexicon. In addition, [6] proposes a framework which discovers acoustic units by clustering together context-dependent grapheme models, and then generates an associated pronunciation lexicon from the initial grapheme-based recognition system. Results on WSJ show the proposed approaches allow for a 13% reduction in WER, and have many implications for low-resourced languages such as the Babel dataset.
Day 3 – ASR in Applications
The third day focused on the impact ASR is making in various applications.
A keynote was delivered on Utilization of ASRU technology - present and future (Joseph Olive), while invited talks were given on the topics of Augmenting conversations with a speech understanding anticipatory search engine (Marsal Gavalda, Expect Labs), Calibration of binary and multiclass probabilistic classifiers in automatic speaker and language recognition (Niko Brummer, Agnitio), Speech technologies for data mining, voice analytics and voice biometry (Petr Schwarz, Phonexia and Brno University of Technology), From the Lab to the Living Room: The Challenges of Building Speech-Driven Applications for Children (Brian Langner, ToyTalk), and The growing role of speech in Google products, by Pedro Moreno.
One interesting paper dealing with spoken language understanding was [7], which looked at a joint model for intent detection and slot filling based on convolutional neural networks (CNN). The proposed architecture shows promising results on a variety of real-world ASR applications. In addition, [8] looks at using linguistic knowledge for query understanding by extracting a set of syntactic structural features and semantic dependency features from query parse trees to enhance inference model learning. Experiments on real natural language queries indicate that using additional linguistic knowledge can improve query understanding results across various real-world tasks.
Day 4 – What’s Wrong with ASR?
Finally, Jordan Cohen and Steve Wegmann jointly delivered a keynote discussing what incorrect assumptions our current modeling approaches make, and what research directions could potentially be pursued to improve ASR performance. For example, HMMs have been around for 40 years, but make poor assumptions such as frame independence. Wegmann also had an interesting paper related to the theme of the day. In this paper [9], Wegmann appies a diagnostic analysis to the performance metric, actual term weighted value (ATWV), used in the Babel task. His analysis looks at the large ATWV gains that often occur due to system combination by increasing the number of true hits, and questions if gains can be obtained without huge expenditures needed by system combination.
Also, for the first time at ASRU, posters were left hanging throughout the conference, and they were being presented in "authors-be-at-their-posters" sessions, which lead to many in-depth discussions also over coffee, during lunch, or in the evening, simply because the poster was still there, to refresh one's memory, or clarify an idea.
References
[1] G. Hinton, L. Deng, D. Yu, G. Dahl, A. Mohamed, N. Jaitly, A. Senior, V. Vanhoucke, P. Nguyen, T. N. Sainath, and B. Kingsbury, “Deep Neural Networks for Acoustic Modeling in Speech Recognition,” IEEE Signal Processing Magazine, vol. 29, no. 6, pp. 82–97, 2012.
[2] G. Saon, H. Soltau, M. Picheny, D. Nahamoo, “Speaker Adaptation of Neural Network Acoustic Models using I-Vectors,” in Proc. ASRU, 2013.
[3] K. Vesely, M. Hannemann, L. Burget, “Semi-supervised Training of Deep Neural Networks,” in Proc. ASRU, 2013.
[4] K. Demuynck and F. Friefenbach, “Porting Concepts from DNNs back to GMMs,” in Proc. ASRU, 2013.
[5] L. Lu, A. Ghoshal and S. Renals, “Acoustic Data-Driven Pronunciation Lexicon for Large Vocabulary Speech Recognition,” in Proc. ASRU, 2013.
[6] W. Hartmann, A. Roy, L. Lamel and J.L. Gauvain, “Acoustic Unit Discovery and Pronunciation Generation from a Grapheme-Based Lexicon,” in Proc. ASRU, 2013.
[7] P. Xu and R. Sarikaya, “Convolutional Neural Network Based Triangular CRF for Joint Intent Detection and Slot Filling,” in Proc. ASRU, 2013.
[8] J. Liu, P. Pasupat, Y. Wang, S. Cyphers, and J. Glass, "Query Understanding Enhanced by Hierarchical Parsing Structures," in Proc. ASRU, 2013.
[9] S. Wegmann, A. Faria, A. Janin, K. Riedhammer, N. Morgan, “The TAO of ATWV: Probabing Mysteries of Keyword Search Performance,” in Proc. ASRU, 2013.
Speech and Audio Highlights from MediaEval 2013
Gareth J. F. Jones, Martha Larson
SLTC Newsletter, February 2014
MediaEval is a benchmarking initiative dedicated to evaluating new algorithms for multimedia access and retrieval. While it emphasizes the 'multi' in multimedia and focuses on human and social aspects of multimedia task, speech and audio processing is a key component of several MediaEval tasks each year. This article overviews the tasks with significant speech and audio elements from the MediaEval 2013 multimedia evaluation benchmark and looks ahead to the MediaEval 2014 campaign [1].
MediaEval 2013 featured a total of 12 tasks exploring aspects of multimedia indexing, search and interaction, five of which involved significant elements of speech and audio processing.
The “Spoken Web Search” (SWS) Task aimed to perform audio search in multiple languages and acoustic conditions where there are very few resources available to develop a solution for each individual language. The operational setting of the task was to imagine that you want to build a simple speech recognition system, or at least a spoken term detection (STD) or keyword spotting (KWS) system in a new dialect, language or acoustic condition, for which only a small number of audio examples are available. The research aimed to explore the question of whether it is possible to do something useful (e.g. identify the topic of a query) by using only the very limited resources available.
The task involved searching for audio content within audio content using an audio query, and contained audio for 9 different languages. Participants were required to build a language independent audio search system so that, given an audio query, it should be able to find the appropriate audio file(s) and the exact location(s) of a query term within these audio file(s). Evaluation was performed using standard NIST metrics and some other indicators. The SWS task at MediaEval 2013 expanded the size of the test dataset and number of languages over similar tasks held in 2011 and 2012. In addition, a baseline system was being offered to first-time participants as a virtual kitchen application [2].
In 2014, the task will continue under the new name of "QUESST": Query by Example Search on Speech.
The Search and Hyperlinking (S&H) Task consisted of two sub-tasks: (i) answering known-item queries from a collection of broadcast TV material, and (ii) automatically linking anchors within the known-item to other parts of the video collection. The S&H Task envisioned the following scenario: a user is searching for a segment of video that they know to be contained in a video collection. If the user finds the segment, she may wish to see further information about some aspect of this segment. This use scenario is a refinement of the previous S&H task at MediaEval 2012.
The dataset for both subtasks was a collection of 1,260 hours of video provided by the BBC. The average length of a video was roughly 30 minutes and most videos were in the English language. The collection was used both for training and testing of systems. Known-items and queries to locate them were created by volunteer subjects in sessions at the BBC offices, and relevant links were identified using crowdsourcing with Amazon Mechanical Turk. The BBC kindly provided human generated textual metadata and manual transcripts for each video. Participants were also provided with the output of two automatic speech recognition (ASR) systems and features created using automatic visual analysis.
The Similar Segments in Social Speech Task was a new task at MediaEval 2013. The task involved finding segments similar to a query segment, in a multimedia collection of informal, unstructured dialogs among members of a small community. The task was motivated by the following scenario. With users’ growing willingness to share personal activity information, the eventual expansion of social media to include social multimedia, such as video and audio recordings of casual interactions, seems inevitable. To unlock the potential value of this material, new methods need to be developed for searching such records. This requires the development of reliable models of the similarity between dialogue-region pairs. The specific motivating task was as follows: A new member has joined an organization or social group that has a small archive of conversations among members. He starts to listen, looking for any information that can help him better understand, participate in, enjoy, find friends in, and succeed in this group. As he listens to the archive (perhaps at random, perhaps based on some social tags, perhaps based on an initial keyword search), he finds something of interest. He marks this region of interest and requests “more like this”. The system returns a set of “jump-in” points, places in the archive to which he could jump and start listening/watching with the expectation of finding something similar.
One dimension of MediaEval’s interest in audio processing is tasks relating to music. MusiClef was a new task at MediaEval 2013, having previously formed part of the CLEF evaluation benchmark [3]. The MusiClef 2013: Soundtrack Selection for Commercials Task aimed at analyzing music usage in TV commercials and determining music that fits a given commercial video. This task is usually carried out by music consultants, who select a song to advertise a particular brand or a given product. By contrast, the MusiClef 2013 task aimed at automating this process by taking into account both context- and content-based information about the video, the brand, and the music.
Music is composed to be emotionally expressive. The Emotion in Music Task was another new task at MediaEval 2013, and sought to develop tools for navigating today’s vast digital music libraries, based on the assumption that emotional associations provide an especially natural domain for indexing and recommendation. Because there are a myriad of challenges to such a task, powerful tools are required for the development of systems that automate the prediction of emotion in music. As such, a considerable amount of work was dedicated to the development of automatic music emotion recognition (MER) systems. The corpus used for this task employed Creative Commons (CC) licensed music from the Free Music Archive4 (FMA), which enabled the content to be redistributed to the participants with annotations created via crowdsourcing using Amazon Mechanical Turk.
Other tasks at MediaEval 2013 included the Placing Task which seeks to place images on a world map, that is, to automatically estimate the latitude/longitude coordinates at which a photograph was taken. The main Placing Task has featured at MediaEval for several year, a newly introduced secondary task for 2013 was Placeability Prediction which asked participants to estimate the error of their predicted location. Annotating images with this kind of geographical location tag, or geotags, has a number of applications in personalization, recommendation, crisis management and archiving. Currently, the vast majority of images online are not labelled with this kind of data. The data for this task was drawn from Flickr. In comparison to previous editions of this task, the test set has not only increased drastically in size, but was also been derived according to different assumptions in order to model a more realistic use-case scenario.
The Violent Scenes Detection Task derives directly from a Technicolor use case which aims at easing a user’s selection process from a movie database, and ran for the third time at MediaEval 2013. This task was to automatically analyse movie content with the objective of identifying violent actions in the content. Another returning task was Social Event Detection (SED) task which required participants to discover social events and organize the related media items in event-specific clusters within a collection of Web multimedia content. Social events are defined as events that are planned by people, attended by people and for which the social multimedia are also captured by people. The Visual Privacy Task (VPT) aimed at exploring how image processing, computer vision and scrambling techniques can deliver technological solutions to some visual privacy problems. The evaluation was performed using both video analytics algorithms and user studies so as to provide both subjective and objective evaluation of privacy protection techniques.
The MediaEval 2013 campaign culminated in a very energetic and successful 2 days workshop in Barcelona, Spain October 2013 attended by 100 task organisers and participants.
The tasks for each MediaEval campaign are chosen following an open call based on results of a public questionnaire exploring the interest of the research community in them. The questionnaire for MediaEval 2014 has recently concluded and selection and details of the tasks to be offered are currently being finalised. Task registration will open in March 2014, details will be available from the MediaEval website [1].
- [1] http://www.multimediaeval.org/
- [2] http://speechkitchen.cse.ohio-state.edu/
- [3] http://www.clef-initiative.eu/
Gareth J. F. Jones and Martha Larson are coordinators of the MediaEval Benchmarking Initiative for Multimedia Evaluation.
Full proceedings of MediaEval 2013 are available from: http://ceur-ws.org/Vol-1043/. More information can be found at http://www.multimediaeval.org/.
Asia Information Retrieval Societies conference 2013
Rafael E. Banchs, Min Zhang, Ming Hui Dong
SLTC Newsletter, February 2014
The Chinese and Oriental Languages Information Processing Society (COLIPS) hosted the ninth edition of the Asia Information Retrieval Societies conference (AIRS 2013) from December 9 to 11, 2013 in Singapore. The Asia Information Retrieval Societies Conference (AIRS) aims to bring together researchers and developers to share new ideas and recent advances in the field of Information Retrieval (IR) and its applications in text, speech, image, multimedia and social data.
Given the increasing adoption of mobile platforms for information services in general, such as social media, information retrieval and e-commerce, the use of speech in the context of the information society is gaining preponderance at a very fast pace. New service paradigms, as well as new business models, are emerging around the use of speech and natural language interfaces. In this sense, AIRS 2013 has provided an excellent environment for technical discussion and debate on the most relevant and current trends for the modern information society, igniting the engines and preparing the Singapore’s scientific landscape for the upcoming INTERSPEECH 2014.
This year, AIRS 2013 received a total of 109 submissions, which were reviewed by a program committee of 143 specialists. After peer reviewing, 45 submissions were accepted for being included in the proceedings and presented at the conference. In total 27 papers were selected for oral presentations and 18 for poster presentations. The proceedings were published as a Springer’s LNCS Series Volume 8281.
The conference was attended by 85 participants from more than 20 countries. About 30% of the registered participants were students. The program also included an Invited Keynote Speech: “Information technologies as innovation drivers in the financial and banking industry” given by Ketan Samani, who is the Executive Director for Regional eBusiness of the Group Consumer Banking Department of DBS Bank in Singapore.
The Best Paper Award was conferred to Laure Soulier, Lynda Tamine and Wahiba Bahsoun for their paper “A Collaborative Document Ranking Model for a Multi-Faceted Search”. In addition, best oral and best poster presentation awards were voted by the participants. The best oral presentation award was conferred to Alistair Moffat, for his paper “Seven Numeric Properties of Effectiveness Metrics”. Finally, due to a tie in the collected votes, the best poster award was conferred to two teams: Qianli Xing, Yiqun Liu, Min Zhang, Shaoping Ma and Kuo Zhang for their paper “Characterizing Expertise of Search Engine Users” and Rajendra Prasath, Aidan Duane and Philip O'Reilly for their paper “Topic Assisted Fusion to Re-Rank Texts for Multi-Faceted Information Retrieval”.
In addition to the technical program, AIRS 2013 featured two main social events: a cocktail session that ran in parallel with the poster presentations at the Rendezvous Grand Hotel Singapore, a modern upscale hotel that is located in the heart of the city; and a conference banquet, which was held at The Halia at Raffles, an urban, casual-chic restaurant and sibling of the award-winning Halia at Singapore Botanic Gardens.
Acknowledgements and more Information
Rafael E. Banchs (rembanchs@i2r.a-star.edu.sg) is a Research Scientist at the HLT department of I2R in Singapore. His research interests are in the areas of Information Retrieval, Machine Translation and Dialogue Systems.
Min Zhang (zhangminmt@hotmail.com) is professor at the School of Computer Science and Technology, Soochow University, Suzhou, China. His research interests are in the areas of Machine Translation and Information Retrieval.
Min Hui Dong (mhdong@i2r.a-star.edu.sg) is vice-president of COLIPS and Research Scientist at the HLT department of I2R in Singapore. His research interests are in the areas of Speech Processing and Synthesis.