Group items tagged

Transcriber is a tool for assisting the manual annotation of speech signals. It provides a user-friendly graphical user interface for segmenting long duration speech recordings, transcribing them, and labeling speech turns, topic changes and acoustic conditions. It is more specifically designed for the annotation of broadcast news recordings, for creating corpora used in the development of automatic broadcast news transcription systems, but its features might be found useful in other areas of speech research.

VoxForge was set up to collect transcribed speech for use in Open Source Speech Recognition Engines ("SRE"s) such as such as ISIP, HTK, Julius and Sphinx. We will categorize and make available all submitted audio files (also called a 'Speech Corpus") and Acoustic Models in GPL format. Users can record themselves reading text excerpts in their native language, or they can access the recordings already recorded in all languages

Evaluation of NLP and Speech Tools for Italian Following the success of Evalita 2007, we are pleased to announce Evalita 2009, the second evaluation campaign of Natural Language Processing tools for Italian, supported by the NLP working group of AI*IA. The general objective of Evalita is to promote the development of language and speech technologies for the Italian language, providing a shared framework where different systems and approaches can be evaluated in a consistent manner. This year in conjunction with AISV (the Italian Association of Speech Science), Evalita will also hold the first speech input technology evaluation for Italian.

""Julius" is a high-performance, two-pass large vocabulary continuous speech recognition (LVCSR) decoder software for speech-related researchers and developers. Based on word N-gram and context-dependent HMM, it can perform almost real-time decoding on most current PCs in 60k word dictation task. Major search techniques are fully incorporated such as tree lexicon, N-gram factoring, cross-word context dependency handling, enveloped beam search, Gaussian pruning, Gaussian selection, etc. Besides search efficiency, it is also modularized carefully to be independent from model structures, and various HMM types are supported such as shared-state triphones and tied-mixture models, with any number of mixtures, states, or phones. Standard formats are adopted to cope with other free modeling toolkit such as HTK, CMU-Cam SLM toolkit, etc. The main platform is Linux and other Unix workstations, and also works on Windows. Most recent version is developed on Linux and Windows (cygwin / mingw), and also has Microsoft SAPI version. Julius is distributed with open license together with source codes. Note: you should prepare a language model and an acoustic model to run a speech recognition with Julius. "

Open source toolkit for speech recognition

A little demo of a program that pronounces words in various languages. A good quick reference and definitely fun to play around with!

ANVILL (A National Virtual Language Lab) is a speech-based toolbox for language teachers. Like the language lab console of old, it's focused on the practice of oral/aural language, but at its core are very modern web-based audio and video tools from duber dot com: Voiceboards (an asynchronous speech tool), LiveChat (a 4-way conversation tool), and Voicerecorder (a widget for instant recordings). Each really opens up the scope and sequence of spoken language tasks. Creating media-rich lessons in ANVILL is simple and straightforward. There are templates for audio, video, and image tasks; there are also tools for text-based discussions like blogs and forums. ANVILL's assessment tool,Quizzes and Surveys, gives teachers an easy way to assign and respond to spoken language tasks. In addition, ANVILL is a course management system. Teachers who are doing distance or hybrid courses have a simple means of managing their students as well as their curriculum. We think you'll find that ANVILL is quite flexible, and permits the kind of extra listening and speaking practice that language students need so much. ANVILL is a project of the Yamada Language Center at the University of Oregon. These tools are currently in use at UO and at a select number of K-12 sites around the Northwest. Thanks to our National Foreign Language Resource Center partners at CASLS, we are able to offer ANVILL to educators at no cost. Here is an overview of its features. If you think it would be helpful to you and your students, contact us to set up an account. Our two-year pilot study has just gotten underway and we're looking for language teachers at all levels to use it and tell us what they think.

VOICE comprises naturally occurring, non-scripted face-to-face interactions in English as a lingua franca (ELF). The recordings made for VOICE are keyboarded by trained transcribers and stored as a computerized corpus. Currently VOICE comprises 1 million words of spoken ELF interactions, equalling approximately 120 hours of transcribed speech. The speakers recorded in VOICE are experienced ELF speakers from a wide range of first language backgrounds. So far, VOICE includes approximately 1250 ELF speakers with approximately 50 different first languages (disregarding varieties of the respective languages). In the initial phase, VOICE focuses mainly, though not exclusively, on European ELF speakers. The ELF interactions recorded cover a range of different speech events in terms of domain (professional, educational, leisure), function (exchanging information, enacting social relationships), and participant roles and relationships (acquainted vs. unacquainted, symmetrical vs. asymmetrical).

"Ipadio technology links up the telephone networks with the internet, enabling the live broadcast of audio directly to the internet… all from a standard phone. There's never been an easier way to record oral assessments, create revision podcasts, or even collect homework. The process involves a simple registration of your phone and the selection of a password. You are then given a toll free number to call, enter the password, and start recording. Your recording is instantly available as a phlog (phone blog) on your own channel at Ipadio. Don't have the cell phone you registered with Ipadio. There is an easy solution, just use any phone and complete the additional step of entering your registered phone number. After the recording you can visit your own channel and listen to the phlog, email it, embed it in a web page, or even download it as an mp3 to be used in your own multimedia application. You even have the opportunity to edit your phlog and make it public. Ipadio even converts the spoken words to text! Imagine the educational uses of that features. There is even an iPhone and Android app."

GhostReader is a multilingual speech solution for Mac OS X that allows you to listen to your documents with naturally sounding voices in a language of choice.

"BrowseAloud reads web pages aloud for people who find it difficult to read online. Reading large amounts of text on screen can be difficult for those with literacy and visual impairments."

A collection of lesson plans and materials for teaching speech acts/pragmatics. Focuses on English, but ideas could be applied to other languages.

KEYNOTE SPEAKERS Friday, April 3: Barbara Sawhill (Oberlin College) Saturday, April 4: Gus Leonard (CSU Monterey Bay) In addition to the keynote speeches, the conference will consist of various workshops, joint presentations and panels centered around the use of technology and language learning. This collaborative conference, co-located with SWALLT, is designed for the expert and novice alike! April 3-5, 2009

Shows a cross-section diagram of the speech organs, and allows you to use buttons to change voicing, nasality, lip position, and tongue position, showing the IPA character for the resulting sound.

"This customizable easy to use toolbar (See the toolbar in action in its video tour) seamlessly integrates with familiar applications including Microsoft Word, Internet Explorer, and Adobe Reader, allowing students to access the support tools they need from within mainstream software programs. It highlights and reads text aloud using natural sounding voices and also includes a comprehensive set of support tools for reading, writing, studying, research, and test taking. Read&Write GOLD allows teachers to provide the differentiated instruction that is necessary to meet the needs of diverse student populations and ensure that all students reach their potential and achieve success. "

"NonVisual Desktop Access (NVDA) is a free and open source screen reader for the Microsoft Windows operating system. Providing feedback via synthetic speech and Braille, it enables blind or vision impaired people to access computers running Windows for no more cost than a sighted person. Major features include support for over 20 languages and the ability to run entirely from a USB drive with no installation."

The Berkeley Language Center has created a library of foreign language clips, each tagged for the vocabulary in the target language and for English descriptors (culture, linguistics, speech acts). Instructors are able to search the library for a clip of interest and then 'order' the clip (a URL is created linking to an mp4 file; the URL stays active for two weeks). Currently we have 5200 clips in 15 languages, and those numbers continue to grow.

This wiki describes tools and formats for creating and managing linguistic annotations. `Linguistic annotation' covers any descriptive or analytic notations applied to raw language data. The basic data may be in the form of time functions -- audio, video and/or physiological recordings -- or it may be textual. The added notations may include transcriptions of all sorts (from phonetic features to discourse structures), part-of-speech and sense tagging, syntactic analysis, "named entity" identification, co-reference annotation, and so on. The focus is on tools which have been widely used for constructing annotated linguistic databases, and on the formats commonly adopted by such tools and databases.

The International Corpus of Learner English (Version 2) is a corpus of writing by higher intermediate to advanced learners of English. It contains 3.7 million words of EFL writing from learners representing 16 different mother tongue backgrounds (Bulgarian, Chinese, Czech, Dutch, Finnish, French, German, Italian, Japanese, Norwegian, Polish, Russian, Spanish, Swedish, Turkish and Tswana). It differs from the first version published in 2002 not only by its increased size and range of learner populations, but also by its interface, which contains two new functionalities: built-in concordancer allowing users to search for word forms, lemmas and/or part-of-speech tags and breakdown of the query results according to the learner profile information.

Submission Deadline: Tuesday, September 15, 2009 Acceptance Notification: Early October Abstracts are invited from undergraduate and graduate students for 15-minute presentations including, but not limited to, relationships between and among language, linguistics, and their many applications: * American Sign Language * natural and artificial languages * extinct and endangered languages * speech pathology and hearing sciences * language and cognition * theoretical and applied linguistics * structural, semantic, or phonological analyses * comparative grammars * discourse/text analysis * applications in rhetoric, the arts, and the humanities * applications in academic, cultural, and sociological constructs