6. Participating Institutions : All academic institutes, research organizations and Corporate R&D groups from India and abroad working on Indian languages will be encouraged to participate in LDC-IL. Besides the lead institutions already mentioned, other possible interested institutes and organizations may include: Different Indian Universities with major departments of Linguistics and computer science/ ISI Calcutta; TIFR Mumbai; HP Labs India; IBM; C-DOT; C-DAC; Tata InfoTech All other IITs; KHS; NCPUL; Rashtriya Sanskrit Sansthan; TDIL, MIT. Departments of Linguistics of various universities in India, to begin with CALTS of the University of Hyderabad. This list is open ended and suggestive and not exclusive.

7. Location : The services under the proposed LDC-IL will be hosted and managed by the Central Institute of Indian Languages, Mysore. This would leverage expertise available in the institution and their prior experience in creating databases. The host institute will require additional space, human resources and other facilities to run the full-time activity of the development and maintenance of the LDC-IL. It is important to mention that all engagements of the personnel for the LDC-IL will be on contractual basis selected through appropriate procedure following the norms fixed by the Government of India for reservation etc. However, as a total R&D activity under the LDC-IL, it is natural that a number of national Institutions will be involved. They will be invited to contribute to the areas listed under Annexures 1 through 4 where specific research areas are identified and elaborated. All these activities and the funding required to carry on such R & D work will be coordinated by the CIIL. However, depending on the experience of the lead institutions, the relevant sub-committees under a project Advisory Committees (PAC) will help in arriving at appropriate decisions.

8. Funding - Grants : For the Core funding, the Department of Secondary and Higher Education will create an appropriate scheme through a substantial one-time grant during Xth Plans, and possibly another set of grants during the next plan to the host institution, i.e. CIIL will be released. As the nodal agency, CIIL will further distribute the relevant funding for specific sub-components of the scheme to other academic institutions based on an intent of understanding to be signed between Heads of such institutions and CIIL. It is also understood that this release of funds and its management will be in a project mode through CIIL's PL account. CIIL may be allowed to set up a separate account for only LDC-IL activities. All receipts and payments will be to this account for potential users of the services and the linguistic data - creating a corpus fund and generating revenues which would make the scheme progress towards self-sustenance in a ten year's time frame. It is needless to mention that an annual accounts as to utilizations of funds as well as revenue generation will be submitted by CIIL to the MHRD along with a progress report in every quarter. Out of the revenue generated 40% will go towards creation of corpus fund and the 60% will be rolled back to towards implementing the tasks.

9. Annual Membership fee :

(a) Open to academic institutes, Research Organizations, and Corporate sector from all over the world.

(b) Members will be encouraged to contribute databases and offered a monetary share of the revenues earned by the databases they contribute

(c) The databases will be available for R&D purposes to all members and non-members on payment of the appropriate fee, with a license for use only. The organization will be asked to sign a License Agreement to the effect that the databases will only be used for research and development and may not be distributed by it to any other institute or individual either free or for a fee. However, the IP and the copyright of any product developed as a result of such an R&D activity shall lie with the organization that has created the product.

Differential rate of annual fee will be charged as given below:

India:

1. Individual Researchers: Rs.2000/- per annum
2. Educational Institutions: Rs.20,000/- per annum
3. Software and related industry : Rs.2,00,000/- per annum

Other countries

1. Individual Researchers: $ 2000/- per annum
2. Educational Institutions: $ 20,000/- per annum
3. Software and related industry : $ 50,000/- per annum

10. Governance of the LDC-IL :

11. About this Proposal : This proposal on LDC-IL has evolved through participation of scholars from different Institutions in India and abroad. Summary of the same is as follows:

(a) On August 13, 2003 ) Prof. Udaya Narayana Singh, Director , Central Institute of Indian Languages made a presentation about the concept of such a consortium and need for creating the same by the Ministry of HRD.

(b) On August 17 and 18, 2003 an International Workshop on Linguistic Data Consortium for Indian Languages was held at the Institute. This was inaugurated by Smt. Kumud Bansal, Additional Secretary, and Smt. Bela Banerjee Joint Secretary (L) from the Ministry also participated. This was conducted in collaboration with IIIT Hyderabad and HPLabs, India, attended by participants from LDC, University of Pennsylvania and most of the Institutes involved in computational linguistics in India.

(c) On August 19, 2003 a follow up meeting was held in the Indian Institute of Science, Bangalore where in a select group of experts from Institutions involved in language technology applications including HPLabs, MAIT participated and formed a group of experts from IITB, IITM, IISc, IIIT Hyderabad with the Director CIIL as Coordinator to formulate a formal proposal to be submitted to the Ministry.

(d) Through teleconference and email correspondences a proposal was formulated and sent to the Ministry vide letter number No.F.2-1/2003/LDC-IL November 18, 2003.

(e) On December 19, 2003 a meeting of representatives of lead Institutes was held at the Institute to discuss the proposal submitted to the Ministry.
(f) With additional inputs received from them, the present proposal is formulated and presented on February 24, 2004 in the MHRD.

(g) Further modified to include suggestions after the presentation by the Director CIIL before the Secretary, Additional Secretary, Joint Secretary(L), Director(Finance) in the Department of Education, Ministry of HRD on February 24, 2004.

(h) On April 16, 2004, the present version of the proposal was presented by the Director, CIIL before the Advisory Committee of the TDIL-MCIT (Technology Development in Indian Languages, Ministry of Communication & Information Technology). The Financial Advisory of the MCIT was also present during the presentation, and the suggestion of the members are included.

(i) Preparation of EFC should start after submission of the final version of the proposal.

(j) Within six months the project should start functioning.

12. Budget

Broad Outline :Rs. 221.60 lakhs per year. Total: Rupees 1772.8 lakhs (variable, please see note below) for the next eight years covering the present and next plan period.

Break up of figures for one year will be as follows :

NOTE:

1. This is a broad indication of proposed expenditure under different heads. The Director CIIL on the advise of the Project Advisory Committee of the LDC-IL may be authorized to make re-appropriation of funds from among the heads indicated above, without exceeding the over all sanctioned budget.
2. The first three years of the project period are incubation years, wherein no heavy flow of income is anticipated from the project.
3. From the third year onwards, flow of income is expected. It is estimated that in the initial period, the annual revenue may be 8% to 10% of the annual investment as projected in the budget proposal given above. If this income is placed in a corpus fund to be used for LDC-IL, in the initial period, there will be a turn over of Rs.17.73 lakhs to Rs.22.16 lakhs per year. From the sixth year of the project, the revenue is expected to be around 25% to 35% of the amount invested, i.e. Rs.55.4 lakhs to Rs.66.48 lakhs annually. Once these are contributed to create a corpus, it is likely that at the end of eight years, the corpus funds will have a figure between Rs. 201.66 lakhs to Rs. 243.76 lakhs plus interests (whatsoever). It is proposed that beyond eight years, the Ministry may consider giving only grants for events (Rs.50 lakhs), tasks of software development (Rs.64.76 lakhs), and maintenance of equipments (Rs.15.24 lakhs), i.e. Rs.130 lakhs a years. The services of the personnel and the IPR costs will be paid from the interests of the corpus funds (Rs.14.63 lakhs) plus further income, i.e. 66.48 lakhs, i.e. Rs.81.11 lakhs generated annually. Added to the maintenance cost of Rs.130 lakhs, the total comes to Rs.211.11 lakhs or so.
4. The project is anticipated to be cent percent self sufficient/financing by the end of Eleventh Plan period or by the end of eight years of its commencement.
5. In case the people in service in the Government or Autonomous Institutions in substantial capacity are selected their service and salary will be protected.

Annexure-1: Speech Recognition and Synthesis

1. Introduction :

The objective of data collection effort is to primarily build speech recognition and synthesis systems for Indian languages. Although there are such ASR and TTS systems available around the world for a number of mainstream languages, commercially viable speech systems for Indian Languages are not available

Voice User Interfaces for IT applications and services have become more and more prevalent for languages like English, and are valued for their ease of access, especially in telephony-based applications. In a country like India, where the majority of the population is not comfortable using English and given the relatively lower rates of literacy, local language speech interfaces can provide access to IT applications and services, through internet and/or telephones, to the masses. If such technology is available in Indian languages, people in various semi-urban and rural parts of India will be able to use telephones and Internet to access a wide range of services and information on health, agriculture, travel, etc. However, for this to become a reality, a computer has to be able to accept speech input in the user's language and provide speech output. Also, in multilingual India, if speech technology is coupled with translation systems between the various Indian languages, services and information can be provided across languages more easily.

Although speech technology has been the focus of research in India for a number of years and the technology itself has matured for real-world applications, the main obstacle in customizing this technology for various Indian languages is the lack of appropriate annotated speech databases in these languages. The focus here is (i) to collect data that can be used for building speech enabled systems in Indian languages and (ii) to develop tools that facilitate collection of high quality speech data.

2. Background - Speech Recognition :

The task of automatic Speech recognition is the task of converting any speech signal into its orthographic representation. There are two different categories of speech recognition systems:

oIsolated word recognition and connected word systems as in command and control applications.
oContinuous speech recognition systems. In continuous speech recognition there are two different categories; read speech and spontaneous speech.

3. Background- Speech Synthesis :

The task of speech synthesis to convert written text (orthographic representation to speech). The vocabulary should not be restricted for speech synthesis and synthesized speech must be close to natural speech. To enable unrestricted speech synthesis, the sentence is normally converted to a sequence of basic units. Then appropriate rules of synthesis are employed to produce speech sounding natural.

The focus is primarily on building (a) vocabulary independent speech to speech translation (for a pair of Indian languages) and (b) vocabulary dependent isolated word recognition in the Indian languages.

4. Long Term Goal:

The grand vision of this project is to collect data to provide speech-to-speech translation from each and every language to each and every other language spoken in India (including Indian English). Such a system would include unlimited vocabulary speech synthesis and recognition systems for every Indian language coupled with machine translation systems between those languages. The block diagram given below describes the basic architecture of such a system.

Speech input
in language A Recognized Text in Language A

5. Short Term Goal:

To create databases for building (a) bi-directional speech to speech translation system of read speech for a pair of Indian languages, namely, Hindi-Telugu, (b) a speech recognition system for Indian English. Further, it is desired to collect large vocabulary isolated data for the 22 Scheduled Indian languages.

Text in Language A

Translated Text in Language B

Speech Output
in Language B

6. Methodology for Short Term Effort:

Methodologies for data collection and development of tools required for the short-term and long-term goals are given below:

Data collection Effort for Automatic Speech Recognition (ASR)

The data collection effort will involve collection of read and spontaneous speech.

Data required:
Read speech corpora for two Indian languages and Indian English.
Channels:
1. Close talking microphone, on a desktop or laptop.
2. Telephone, both landline and mobile .
Annotation:
The data will be annotated at phoneme, syllable, word and sentence levels.

Data Collection for Isolated Speech Recognition

Channels:
1. Close talking microphone, on a desktop or laptop
2. Telephone, both landline and mobile
Demography:
10,000 words from 300 speakers (150 male, 150 female)

Data Collection for Text to Speech Synthesis

Data Required:
Data will be collected in the form of read-out phonetically balanced text which will ensure coverage of all speech sounds of the language concerned in different prosodic and phonological contexts. The phonetically balanced text will be extracted from a huge text corpus.
Channels:
Speech Synthesis requires high quality recording in an anechoic chamber using high quality microphones and recording equipment.
Demography:
6 speakers: 3 males and 3 females per language.
Annotation:
Data to be annotated at phone, phoneme, syllable, word, and phrase level.

Tools Required for Data Collection and Annotation for ASR and TTS:
Standardization of tools for data collection and annotation is required. Number of different tools are available for annotation of speech data like EMULAB and PRAAT. Convergence on representation, annotation and storage format is required. However, the project will also focus on providing converters across different formats.

Tools for Speech Recognition:
The data will be annotated at phoneme, syllable, word and sentence levels. Tools need to be developed for semiautomatic annotation of speech data. These tools will also be useful for annotating speech synthesis databases. One could adopt LDC's recording of data in the NIST format. This format is comprehensive in that it contains ALL the information about the recording environment, speaker information, sampling rate, number of channels, number of bits/sample, etc.

Tools For Speech Synthesis:
Other than tools for annotating speech databases, text annotation is also required for speech synthesis:
Text Annotation Tools Required:
1. PoS taggers, phrase boundary markers and intonation markers.
2. Identification and standardization of feature vectors for speech synthesis , vocalic/non- vocalic, pause break, distance from a pause break, duration of break, characteristics of the intonation across the entire sound unit.

7. Applications:

oSpeech to Speech translation for a pair of Indian languages, namely, Hindi and Telugu.
oCommand and control applications.
oMultimodal interfaces to the computer in Indian languages.
o E-mail readers over the telephone,
o Readers for the visually disadvantaged.
o Speech enabled Office Suite.

The effort for both Speech Recognition and Speech Synthesis will be repeated across all 22 Scheduled languages. For Speech Recognition, spontaneous speech data will be collected along with read speech. For speech synthesis, data will be collected from professional speakers, with very good voice quality. Additional speech data will be collected to come out with models for prosody (intonation, duration, etc.) to improve the naturalness of synthesized speech. A database (lexicon) of proper names (of Indian origin) will be created, with the equivalent phonetic representation for each of the names.