Welcome to the Summer 2014 edition of the IEEE Speech and Language Processing Technical Committee's Newsletter! This issue of the newsletter includes 7 articles and announcements from 13 contributors, including our own staff reporters and editors. Thank you all for your contributions! This issue includes news about IEEE journals and recent workshops, SLTC call for nominations, and individual contributions.
We believe the newsletter is an ideal forum for updates, reports, announcements and editorials which don't fit well with traditional journals. We welcome your contributions, as well as calls for papers, job announcements, comments and suggestions.
To subscribe to the Newsletter, send an email with the command "subscribe speechnewsdist" in the message body to listserv [at] listserv (dot) ieee [dot] org.
Florian Metze, Editor-in-chief
William Campbell, Editor
Haizhou Li, Editor
Patrick Nguyen, Editor
Douglas O'Shaughnessy
My term as editor-in-chief for the SLTC newsletter is over after two years. Serving for the newsletter has been an energizing experience for me.
Singapore’s rich linguistic landscape, carved by multicultural immigrants, its colonial heritage, and vibrant economic opportunities, makes Singapore a unique place to celebrate the diversity of spoken languages. With the arrival of INTERSPEECH held in Singapore in September 2014, we hope to offer you a taste of Singapore’s colorful multilingual landscape by contributing this news article.
The Member Election Subcommittee of the SLTC is seeking nominations for new SLTC Members. Nominations should be submitted to the Chair of the Member Election Subcommittee.
The 39th International Conference on Speech and Audio Processing was recently hosted in Florence, Italy from May 4-9th, 2014. In this article, we highlight upward and downward trends in the speech community, and compare this to last year’s ICASSP.
The 2014 IEEE Spoken Language Technology Workshop (SLT 2014) will be held in South Lake Tahoe, California and Nevada, on Dec 7-10, 2014. (http://www.slt2014.org/). The main theme of the workshop will be "machine learning in spoken language technologies".
This is a special issue dedicated to continuous space and related methods in natural language processing to mark the merger of IEEE Transactions on Audio, Speech and Language Processing and ACM Transactions on Speech and Language Processing. IEEE/ACM T-ASL features a brand new EDICS for Human Language Technology which reflects the growing interest of Natural Language Processing (NLP) research, recognizing the strong interactions between NLP and speech processing.
The Organizing Committee of INTERSPEECH 2015 is now inviting proposals for satellite workshops, which will be held in proximity to the main conference.
SLTC Newsletter, May 2014
Welcome to the spring SLTC Newsletter. As we are now fresh from an invigorating and productive ICASSP meeting in Florence, allow me to suggest a way to participate more in our Society’s activities: consider submitting a nomination for you or a colleague to our technical committee, the SLTC. As you may know, the SLTC consists of 56 researchers in the speech and language fields. We have direct responsibility for the evaluation of ICASSP paper submissions in the speech and language areas. We also handle all details for our ASRU (Automatic Speech Recognition and Understanding) and SLT (Spoken Language Technology) workshops. We select the locations for these workshops, give advice on many other subjects related to speech and language, and publish the newsletter that you are now reading. (Look for an announcement soon about where ASRU-2015 will take place.)
This September, we will have an election to replace the 18 members whose three-year terms end this December. Details can be found elsewhere in this SLTC newsletter. We are looking for accomplished people in all areas of speech and language processing. The main tasks for each of our members are reviewing approximately 11 paper submissions to ICASSP each year and helping in one of our subcommittees: Language Processing, Electronic Newsletter, Fellows, Workshops, EDICS, Policies and Procedures, Education, Nominations and Awards, Communications, Industry, External Relations, Student Awards, Member Election, Meetings, and Area Chairs. If you have needs in any of these areas, please contact one of our subcommittee members, as listed on our web page: http://www.signalprocessingsociety.org/technical-committees/list/sl-tc/ We look forward with anticipation to the upcoming IEEE SLT Workshop, to be held at held at Harvey’s Lake Tahoe Hotel in Lake Tahoe, Nevada (Dec. 7-10, 2014); http://www.slt2014.org/ is the website. Paper submissions are due by July 21st. The main theme of the workshop will be machine learning in spoken language technologies. The workshop goals include increasing both intra- and inter-community interaction, through keynote/guest speakers from the machine language community, online panel discussions, miniSIGs (special interest groups), and highlight sessions. We especially hope this year to report more on these emerging areas: large-scale spoken language understanding, massive data resources for SLT, unsupervised methods in SLT, capturing and representing world knowledge in SLT, web search with SLT, SLT in social networks, multimedia applications, and Intelligent environments.
As our Technical Committee (SLTC) is committed to supporting the diversity of speech and language topics, as well as reaching out to new communities, I encourage all to consider attending the upcoming Interspeech-2014 conference (http://interspeech2014.org/), which will be held in Singapore, Sept. 14-18 (as a side benefit, Singapore hosts the international F1 auto race immediately after the conference).
In closing, please consider participating at SLT this year in Tahoe and at next year’s ICASSP in Brisbane. We look forward to meeting friends and colleagues at these exciting locations.
Best wishes,
Douglas O'Shaughnessy
Douglas O'Shaughnessy is the Chair of the Speech and Language Processing Technical Committee.
SLTC Newsletter, May 2014
My term as editor-in-chief for the SLTC newsletter is over after two years. Serving for the newsletter has been an energizing experience for me.
First of all, I’d like to thank the great team that I’ve worked with: editors William Campbell, Martin Russell, Haizhou Li and Patrick Nguyen and volunteer senior reporters: Tara N. Sainath, Svetlana Stoyanchev, Antonio Roque, Matthew Marge, Nancy Chen and Navid Shokouhi who contributed to many issues with articles and helped other contributors with editing, SLTC chairs Douglas O’Shaughnessy and John Hansen for their timely contributions, IEEE website person Rupal Bhatt, and all the guest contributors.
During this time period, in addition to continuing the quality of the newsletter as transferred by the previous editors Jason Williams and Mike Seltzer, I focused on recruiting overview articles in areas of high interest to the SLTC community, such as situated conversational interactions, speaker identification, and deep learning, summaries from conference and workshops that aim to give insights to non-participants and refresh the memory of participants, and highlights and summaries of shared task evaluations before and after the events, with the aims of awareness and visibility.
I’m happy to transfer the role to Florian Metze and the team, who already published a great issue! I wish them the best!
Dilek Hakkani-Tür is a senior researcher at Microsoft Research. Her research interests include learning methods for speech and language processing and conversational systems.
SLTC Newsletter, May 2014
Singapore’s rich linguistic landscape, carved by multicultural immigrants, its colonial heritage, and vibrant economic opportunities, makes Singapore a unique place to celebrate the diversity of spoken languages. With the arrival of INTERSPEECH held in Singapore in September 2014, we hope to offer you a taste of Singapore’s colorful multilingual landscape by contributing this news article.
Singapore has four official languages: English, Malay, Mandarin Chinese, and Tamil. Yet the languages in Singapore go way beyond just these four. Examples include Chinese languages such as Hokkien (Min Nan), Teochew, Cantonese, and Hakka; Indian languages such as Hindu and Punjabi; and creoles such as Singlish (colloquial Singaporean English), Baba Malay (Hokkien influenced Malay), and Kristang (Malay influenced Portugese). Below we attempt to walk you through the development of multiligualism in Singapore through introducing some classic examples of Singaporean languages and discussing the corresponding historical context, cultural significance, and language policies.
The documented history of Singapore started as the first settlements were established in the 13th century AD [1]. Singapore was part of different kingdoms and sultanas until the 19th century, when modern Singapore was founded under the impulsion of the British Empire. In 1819, Sir Thomas Stamford Raffles landed in Singapore and established a treaty with the local rulers to develop a new trading station. From this date, the commercial importance of Singapore continuously grew under the influence of Sir Raffles. Singapore remained under British administration until 1963. Singapore was then part of the Federation of Malaya before becoming independent in 1965 and joined the Commonwealth of Nations in 1966.
English has a special status in Singapore. It is the only national language that is not a mother tongue [2]. English is referred to as “cultureless” as it is “disassociated from Western culture” in the Singaporean context [3]. This cultural voiding makes English an ethnically neutral language used as the lingua franca among ethnic groups [4] after replacing the local Malay language [5]. English is the only compulsory language of education. In schools, English is the only first language, as opposed to the second language status delegated to the other official languages (i.e., English is the language used to teach all subjects, including other languages). By promoting the use of English as the official working language, the intention is to not advantage or disadvantage any ethnic group. English is the default language used in all government related affairs, including legislation and parliamentary business.
Standard Singapore English is almost identical to British English or Received Pronunciation [6] though a new standard of pronunciation has been emerging recently [7]. Interestingly, some aspects of the pronunciation of Singapore English cannot be predicted from British English or any other variety of English.
Legend has it that a Prince from Palembang (South Sumatra, Indonesia), descendant of Alexander the Great, named the city Singapura (which means Lion City in Malay) after seeing a beast on the islands in 1299 BCE. While there is no scientific evidence that there were any lions that ever lived in Singapore outside the zoo, the name Singapore is long-lived till today.
Malays are the predominant ethnic group inhabiting the Malay Peninsula, Eastern Sumatra, Brunei, coastal Borneo, part of Thailand, and the Southern Burmese coast. In Singapore’s constitution, Malays are acknowledged as the “indigenous people of Singapore”. Therefore, even though the Malay population in Singapore is only 13%, Malay (Bahasa Melayu) has a special status among the four official languages of Singapore. For example, the national anthem is in Malay to honor the long history of Malay inhabitants.
With almost 220 million of speakers, the Malay language in its various forms unites the fifth largest language community in the world [8]. Origins of Malay language can be traced amongst the very first Austronesian languages, back to 2000 BCE. Through the centuries, the major Indian religions brought a number of Sanskrit and Persian words to the Malay vocabulary while Islamization of Southeast Asia added Arabic influences. Later on, languages from the colonization powers (mainly Dutch and British) and immigrants (Chinese and Tamil) contributed to the diversity of Malay influences. In return Malay words have been loaned in other languages such as English, e.g., rice paddy (from padi, referring to the rice plant Oryza sativa), Orangutan (literal translation: people from the jungle), gecko, compound (enclosed group of buildings, originating from kampong meaning village in Malay), and Rambutan.
Phonologically speaking, Malay is different from many Asian languages and English. Different from many Southeast Asian languages such as Thai and Vietnamese, Malay is not tonal. There are only six vowels in Malay, and they follow the rule of vowel harmony in word constructions. Unlike English, the pronunciation of Malay words is fairly orthographic. Morphologically, Malay is an agglutinative language, where new words can be formed by three methods: (1) affixation: attaching affixes onto a root word, (2) composition: formation of a compound word, (3) reduplication: repetition of words or portions of words. Similar to modern English and Chinese languages, Malay does not use grammatical gender like many European languages. Verbs are not inflected for person or number. Nor are verbs marked for tense either.
Three quarters of the population in Singapore are ethnically Chinese. Dating back to the 10th century, Chinese immigrants have already been settling in Southeast Asia (Indonesian archipelago, Peninsular Malaysia, and Singapore). They were joined by larger numbers of Chinese traders in the 15th through 17th centuries [9]. Most of these immigrants were male, whom married local Malay women. Their descendants are often referred to as Peranakan Chinese or Baba-Nyonya. Their language, Baba Malay, is a creole dialect of the Malay language, which contains many Hokkien words. While a dying language with its contemporary use is limited to the older generation, the Peranakan Chinese culture has maintained. The most well-known examples include the celebration of the Lunar New Year and Nyonya cuisine, which is the blending of Chinese ingredients with various distinct spices and cooking techniques used by the Malay/Indonesian community. While the Peranakan Chinese community is no longer the majority of the ethnically Chinese population in Singapore, many famous Singaporeans are Perankan Chinese, including Lee Kuan Yew, the founding father of modern Singapore.
Nowadays, the majority of the ethnically Chinese population in Singapore are 3rd generation immigrants from Southern China (e.g., Fujian, Guangdong, and Hainan provinces). Recent immigrants from People’s Republic of China (PRC) in the last couple of decades are no longer only from Southern China, adding more linguistic and cultural variety within the ethnically Chinese community in Singapore.
Singapore Mandarin has now replaced Hokkien as the lingua franca of the ethnically Chinese community in Singapore [10]. In the past, the majority Southern-Chinese immigration to Singapore made Hokkien the de facto common language (in 1957, 97% of Chinese Singaporean spoke a non-Mandarin Chinese language at home). The replacement of Hokkien by Mandarin was induced by the Government’s “Speak Mandarin” campaign starting in the 1979. Today, Mandarin is spoken in 47% of the Chinese households while usage of other Chinese languages dropped to 19%. Nevertheless, a large majority of Chinese Singaporeans speaks or understands other Chinese languages, another living example of multilingualism in Singapore.
Singaporean standard Mandarin Chinese, one the four official languages of Singapore, is almost similar to its continental twin: PRC's Putonghua (普通话). Like other varieties of Chinese, it is an analytic language that follows a subject-verb-object structure. The acoustic unit is the syllable (denoted by a Chinese character) consists of an initial and a final (or sometimes just a final), which is encoded by a lexical tone. Mandarin Chinese consists of five tones (including a neutral tone) [11, 12], while other Chinese languages like Hokkien and Cantonese have a more sophisticated tonal system that consists of up to 9 different lexical tones.
Due to the diverse backgrounds of ethnically Chinese immigrants, their blending with the local Malay languages, the English influence of the British colonial era, and the adoption of Mandarin as one of the four official languages, colloquial Singaporean Mandarin is a lively example of multilingualism. For example, basa means market, and originates from bazzar in Malay; deshi (德士) originates from taxi in English; kiasu (a characteristic of Singaporean culture that emphasizes one would over-prepare to ensure he succeeds) is adopted from Hokkien, which literally means afraid to lose. Singapore Mandarin has also a distinctive accent. For example, the neutral fifth tone of Singapore Mandarin is replaced by a distinctive ru tone due to the influence of Southern Chinese languages such as Hokkien, Teochew or Cantonese [13].
Around 9% of the Singaporean population is ethnically Indian, making up the third largest ethnic group in Singapore. The origins of Singaporean-Indians are diverse: usually locally born, they are second, third, fourth or even fifth generation descendants of Punjabi, Hindi, Sindhi and Gujarati-speaking migrants from the Northern India and Malayalees, Telugu, and Tamil-speaking migrants from the Southern India. This latter group is the core of Singaporean-Indian population, which consists 58% of the Indian community.
Historically speaking, Indian immigrants were brought to Singapore during the British colonization to fulfill functions such as clerks, soldiers, traders and English teachers. After 1980, the Singaporean immigration policy aimed at attracting educated people, which also contributed to Indian immigration. In addition to this residential population, many ethnic Indian migrant workers temporarily come to work in Singapore from Bangladesh, Sri Lanka, Malaysia and India) [14].
Over half of the Indian population has Tamil origins. Tamil is one of the longest surviving classical languages in the world. For over 2,000 years, Tamil has a rich literature history and was the first Indian language to be declared a classical language by the India Government. Earliest records of written Tamil were dated from around the 2nd century BC and, despite the significant amount of grammatical and syntactical change, this language demonstrates grammatical continuity across 2 millennia. Major Tamil speaking population can be found in Malaysia, Philippines, Mauritius, South Africa, Indonesia, Thailand, Burma, Reunion and Vietnam. Significant communities can also be found in Canada, England, Fiji, Germany, Netherlands and the United States. It is the official language in Indian states of Tamil Nadu, Puducherry, Andaman and Nicobar Islands as well as in Sri Lanka and Singapore.
Like Malay, Tamil is agglutinative. Affixes are added to words to mark noun class, number, case or verb tense, person, number, mood and voice. Like Finish, Tamil sets no limit to the length and extent of agglutination. This leads to long words with a large number of affixes in which its translation might require several sentences in other languages. Phonology of Tamil is characterized by the use of retroflex consonants and multiple rhotics. Fricatives are only used for foreign phones and do not exist in the original language. Native grammarians classify phonemes into vowels, consonants and a secondary character called āytam, an allophone of /r/ or /s/ at the end of an utterance. Unlike most other Indian languages, aspirated and unaspirated consonants are not distinguished in Tamil, but assigned depending on the position of the consonant in the word. Among all the Indian residents in Singapore, 39% speak Tamil, 39% speak English, 11% speak Malay, and the remaining 11% speak other Indian languages [15]. Tamil is one the two Indian languages taught as a second language in public schools, together with Hindi. It also used in daily newspapers, television, radio channels, and theaters.
Kristang (aka Papia Kristang or Christao) is a Portuguese-based creole language influenced by Malay, English and other languages spoken on the Malaya peninsula. This creole originated from Malacca in 1511, when the Portuguese explorer, Alfonso de Albuquerque, conquered the city for trading rare spices. In order to ensure the loyalty of the local population and to provide manpower, Alfonso de Albuquerque encouraged marriages between Portuguese men and Malay women. In 1641, Portuguese lost Malacca to the Dutch. Dutch men married local “Portuguese” women and embraced their Catholic faith. This mix of Malay, Portuguese and Dutch was known as the Malacca-Portuguese or Jenti Kristang speech community. The Kristang community also settled into other Southeast Asian regions such as Singapore.
Although Kristang has no written form and has never been formally taught in school, it has been passed down from generation to generation, through daily usage and church services. A first proposal for standard orthography was made in late the 1980's by Alan Baxter using Malay orthography. In the 1990s, Joan Marbeck's book "Ungua Andanza" was published, with a “Luso-Malay” orthography. The grammatical structure of Kristang is very close to Malay but a large part of the vocabulary (~95%) is Portuguese, so Kristang is generally quite recognizable to speakers of European Portuguese though many word are considered archaic. Perhaps because of cultural exchanges along trade routes, Kristang has a lot of similarities with other Portuguese-based creoles spoken in Indonesia and East-Timor.
Nowadays, Kristang has 5,000 speakers in Malacca and 400 in Singapore. It is also spoken in some parts of Indonesia and in Australia (region of Perth) due to migrations. Kristang is considered the “last vital variety of a group of East and Southeast Asian Creole Portuguese languages” [6] and categorized as one of the endangered languages in Malaysia. In order to revitalize the language, publications of dictionaries, phrase-books, and language documentation efforts are encouraged. Social media are also to promote the use of Kristang with Facebook pages such as “Yo Falah Linggu Kristang” (I speak the Kristang language). Although most people, including Singaporeans, Malaysians, Portuguese and Dutch, are unaware of its existence, Kristang reminds us of the reminisce of Euroasian heritage in Southeast Asia.
Singlish (colloquial Singapore English) is a vivid and colorful creole example of how languages and speakers interact and mingle: while Singlish is a language variety of English, it is interleaved with slangs from languages such as Hokkien, Teochew, Cantonese, Malay and Tamil, and heavily influenced by Chinese grammar, phonology, and prosody. The complexity of Singlish exemplifies the challenges speech researchers face in developing spoken language technologies to automatically identify, transcribe, and parse colloquial and conversational speech.
Singlish is semi-tonal, as all words of Chinese origin retain their original words, while original English words as well as Malay and Tamil words are non-tonal. In addition, although most varieties of English are stressed-time, Singlish is syllable-timed, giving Singlish a rather staccato feel.
Singlish phonology is primarily British based, with influence of Chinese phonology. For example, the dental fricatives /θ/ and /ð/ are sometimes merged with /t/ and /d/ in certain contexts, so that three sounds like tree and then sounds like den [16]. The voiceless stops /p/, /t/, /k/ are also sometimes unaspirated as in Chinese languages [18]. There is generally no distinction between the non-close front monophthongs, so pet and pat are pronounced the same /pɛt/ [17].
At the vocabulary level, there is often inter-mixing of multiple languages. For example, damn shiok is a slang blending English and Punjabi, used to express extreme pleasure or satisfaction, often in the context of food. Mixing of languages is also reflected from location names. For example, Toa Payoh (literal translation: big swamp), a central district in Singapore, mixes Hokkien and Malay (toa is big in Hokkien and payoh is swamp in Malay). Reduplication is also used in Singlish, which is influenced by Chinese and Malay. Adjectives of one or two syllables can also be repeated for intensification. For example, “You go take the small-small one ah.” (Retrieve the smaller item, please.) The frequent use of already (pronounced more like oreddy) in Singapore English is probably a direct influence of the Hokkien liao particle [18]. For example, “Aiyah, cannot wait any more, must go oreddy.” (Oh dear, I cannot wait any longer. I must leave immediately.)
Singlish is topic-prominent like Chinese and Japanese, meaning that Singlish sentences often begin with a topic followed by a comment of new information. For example, ”Dis country weather very hot one.” (In this country, the weather is very warm.) The topic can be omitted when the context is clear, resulting in constructions that appear to be missing a subject. For example, “No good lah” (This isn’t good.)
Singlish is also known for its colorful usage of interjections from Chinese and Malay influence (examples in previous sentences examples include ah, aiyah, lah). ”lah” is probably the most famous one and a stereotypical interjection which appears to be ubiquitous to non-native speakers of Singlish. It may originate from the Hokkien character (啦), though its usage in Singapore is also influenced by its occurrence in Malay [19]. “lah” has many different usages. It is often used to soften ones tone. For example, “Cannot lah”, “Just drink lah”. It can also be used to indicate impatience with a low tone; e.g., “Eh, hurry up lah!” It can also be used for reassurance: Okay lah. (It's all right. Don't worry about it.) Yet, it can also be used to curse people. For example, “Go and die lah”.
Although Singlish is typically not used in official settings (e.g. school lectures and mainstream media generally use Standard Singapore English), Singlish is quite prevalent in day-to-day interactions with peers, siblings, parents, and elders. It is an effective means to establish rapport (for example, during military service) or for humorous effects for TV and radio shows. From a linguistic perspective, Singlish is a living example of multilingualism in Singapore blending the East and the West.
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.
Anthony Larcher is a scientist in the department of Human Language technology of the Institute for Infocomm Research in Singapore. His research interests include text-dependent and –independent speaker verification as well as language recognition. More information on http://www1.i2r.a-star.edu.sg/~alarcher
SLTC Newsletter, May 2014
The Member Election Subcommittee of the SLTC is seeking nominations for new SLTC Members. Nominations should be submitted to the Chair of the Member Election Subcommittee at larry.heck@ieee.org, with cc to bhiksha@cs.cmu.edu, Fabrice.Lefevre@univ-avignon.fr, and stolcke@icsi.berkeley.edu. Please provide the name, contact information and biography with the nomination.
New member candidates can be self-nominated or nominated by current SLTC members. Members will serve a term of three years. Past SLTC members are eligible to be nominated for a second term. Current SLTC members may also be nominated for a second consecutive term, but would then not be eligible to vote in this new member election. Additional terms are allowed after leaving the SLTC, but at least a 3-year gap in service is required.
New members must be willing to review papers that are submitted to the Society's conferences within the area of the SLTC, review papers for workshops owned or co-owned by the SLTC, serve in the subcommittees established by the SLTC, and perform other duties as assigned. SLTC members will be elected by the members of the SLTC itself based on the needs of the SLTC. The election results will be finalized by 15 November 2014.
Please use the nomination form outlined below and submit a completed nomination form no more than one page in length:
The rights and duties of an SLTC member are described here: http://www.signalprocessingsociety.org/technical-committees/list/sl-tc/
The following are nominee requirements:
The nomination deadline is set as 10 September 2014.
SLTC Newsletter, May 2014
The 39th International Conference on Speech and Audio Processing was recently hosted in Florence, Italy from May 4-9th, 2014. In this article, we highlight upward and downward trends in the speech community, and compare this to last year’s ICASSP.
Deep Neural Networks continue to remain a hot topic in the ASR space, particularly for acoustic modeling. In the acoustic modeling space, last year's ICASSP had just two oral sessions with 12 papers devoted to Neural Network acoustic models, while this year there were two poster sessions and two oral sessions, totaling over 40 papers. Many of the papers focused particularly on improving training time of the networks.
Given the popularity of DNNs in the acoustic modeling space, there has been an increased interest in exploring DNNs and their variants in other areas of ASR such as Robust Speech Recognition, Language Modeling, Speech Synthesis and Speaker Identification. While at least year's ICASSP there were roughly 12 papers in this area, this time we saw over 20 in this area. Given the popularity of the IARPA Babel project and performing keyword search for low-resourced languages, we see increased popularity in Spoken Term Detection. This year, there were two poster sessions, totaling over 28 papers. However, at last year’s ICASSP there was just one oral session with 6 papers.
Also, given the DARPA RATS program, there seems to be continued interest in Robust Speech Recognition. Last year there were two poster sessions and one oral session totaling roughly 30 papers, and this year there were two poster sessions, also totaling roughly 30 papers. Topics in speech synthesis also seem to be more focused this year, with session topics including Stochastic, Statistical and Parametric Speech synthesis. Last year, the topics for speech synthesis were more general and including prosodic methods, HMM-based methods and general Speech synthesis topics. However, there were roughly 30 papers total in both of these areas. Papers in the Spoken Language Understanding and Dialog systems also continue to remain popular. Last year there were two sessions in this area of roughly 15 papers and this year the trend in similar.
In the acoustic modeling area, last year's ICASSP had sessions on discriminative training methods and alternative acoustic modeling techniques such as Gaussian Mixture Models, Kernel Methods and exemplar-based methods. We see fewer and fewer papers in this area, and no main session in acoustic modeling devoted to anything outside of DNNs. We as a community as a whole should maybe start exploring alternative methods that could achieve similar performance without the training time bottlenecks.
Another downward trend is in the Language Modeling space. Last year in this area there was one oral and one poster session, totaling roughly 15 papers. At this year’s ICASSP we see just one poster session with 10 papers.
Tara N. Sainath is a staff writer for the SLTC Newsletter. She is a research scientist at IBM.
SLTC Newsletter, May 2014
The Fifth IEEE Workshop on Spoken Language Technology (SLT 2014) will be held in South Lake Tahoe, Nevada, on Dec 7-10, 2014.
The main theme of the workshop will be "machine learning in spoken language technologies". There will be keynote/guest speakers from the machine learning community. One of the workshop goals is to increase both intra and inter community interactions. Towards this goal, in addition to tutorials and keynote speeches on main workshop theme and emerging areas, this year's SLT will host special sessions and self-organizing Special Interest Group (SIG) meetings, as well as panel discussions before/during workshop. If you want to excite the community about a topic and to have an impact on the workshop content, now this is your chance!
In addition to submitting papers and/or proposing/organizing SIG meetings, you can get involved in workshop organization in different ways: by nominating keynote speakers (nominations@slt2014.org), or by volunteering to be part of workshop organization (volunteers@slt2014.org). Please visit www.slt2014.org for more details.
Submission of papers in all areas of spoken language technology is encouraged, with emphasis on the following topics, including both traditional SLT areas as well as emerging ones:
Traditional Topic Coverage: | speech recognition and synthesis, spoken language understanding, spoken dialog systems, spoken document summarization, machine translation for speech, question answering from speech, speech data mining, spoken document retrieval, spoken language databases, speaker/language recognition, multimodal processing, human/computer interaction, assistive technologies, natural language processing, educational and healthcare applications. |
Emerging Areas: | large scale spoken language understanding, massive data resources for SLT, unsupervised methods in SLT, capturing and representing world knowledge in SLT, web search with SLT, SLT in social networks, multimedia applications, intelligent environments. |
Prospective authors are invited to submit full-length, 4-6 page papers, including figures and references, to the SLT 2014 website (www.slt2014.org).
Special Session (SS) proposal submission: | June 6, 2014 |
Paper Submission: | July 21, 2014 |
Notification of acceptance: | September 5, 2014 |
Demo submission: | September 10, 2014 |
Notification of SS proposals (1st/2nd decision): | June 15/ September 19, 2014 |
Notification of Demo acceptance: | October 10, 2014 |
Early registration deadline: | October 17, 2014 |
Special Interest Group (SIG) proposal submission: | November 21, 2014 |
Workshop Date: | December 7-10, 2014 |
Supported by:
SLTC Newsletter, May 2014
This is a special issue dedicated to continuous space and related methods in natural language processing to mark the merger of IEEE Transactions on Audio, Speech and Language Processing and ACM Transactions on Speech and Language Processing. IEEE/ACM T-ASL features a brand new EDICS for Human Language Technology which reflects the growing interest of Natural Language Processing (NLP) research, recognizing the strong interactions between NLP and speech processing.
Natural Language Processing aims to analyze, understand, and generate languages that humans use naturally. Significant progress in NLP has been achieved in recent years, addressing important and practical real-world problems, enabling mass deployment of large-scale systems. New machine learning paradigms such as deep learning and continuous space methods have contributed to inferring language patterns from increasingly large real-world data and to making predictions about new data more accurate.
One of the challenges in NLP is to represent language in a form that can be processed effectively by computing algorithms. Words in sequence are traditionally treated as discrete symbols, which has its advantages and limitations. The research on continuous space methods provides a promising alternative that describes words and their semantic and syntactic relationships in a different way. In continuous space language modeling, we represent words with real-valued vectors. In this way, conditional probability distributions of words can be learned and expressed as smooth functions of these vectors; similar words are therefore described as neighbors in a continuous space. A Neural Network Language Model is a typical example of such continuous space methods.
Building on the success of acoustic and statistical language modeling, research on artificial (deep) neural networks and continuous space models in general has seen significant progress in mitigating data sparseness, incorporating longer contexts, and modeling morphological, syntactic and semantic relationships across words. As a result, continuous space models are now embedded in many state-of-the-art speech recognition and machine translation systems. This special issue provides a forum to discuss the latest findings on research problems related to the application of continuous space and related models in NLP. We invite papers on various NLP topics, including but not limited to
The authors are required to follow the Author’s Guide for manuscript submission to the IEEE /ACM Transactions on Audio, Speech, and Language Processing at http://www.signalprocessingsociety.org/publications/periodicals/taslp/taslp-author-information
Haizhou Li, Marcello Federico, Xiaodong He, Helen Meng, Isabel Trancoso are the Guest Editors of this special issue.
SLTC Newsletter, May 2014
INTERSPEECH 2015 will be held in the beautiful city of Dresden, Germany, on September 6-10, 2015. The theme is "Speech beyond Speech - Towards a Better Understanding of the Most Important Biosignal".
The Organizing Committee of INTERSPEECH 2015 is now inviting proposals for satellite workshops, which will be held in proximity to the main conference.
The Organizing Committee will work to facilitate the organization of such satellite workshops, to stimulate discussion in research areas related to speech and language, at locations in Central Europe, and around the same time as INTERSPEECH. We are particularly looking forward to proposals from neighboring countries. If you are interested in organizing a satellite workshop, or would like a planned event to be listed as an official satellite event, please contact the organizers or the Satellite Workshops Chair, Florian Metze.
The Satellite Workshop coordinator along with the INTERSPEECH team will help to connect (potential) workshop organizers with local contacts in Germany, if needed, and will try to be helpful with logistics such as payment, publicity, and coordination with ISCA or other events.
Proposals should include:
Proposals for satellite workshops should be submitted by email to Florian Metze by August 31st, 2014.
We strongly recommend that organizers also apply for ISCA approval/ sponsorship, which will greatly facilitate acceptance as an INTERSPEECH satellite event. We plan to notify proposers no later than October 30, 2014.
If you have any questions about whether a potential event would be a good candidate for an INTERSPEECH 2015 satellite workshop feel free to contact the INTERSPEECH 2015 Satellite Workshops Chair.
Sincerely,
Florian Metze
Florian Metze is the Satellite Workshops Chair of INTERSPEECH 2015