Alveo Virtual Laboratory
The Alveo Virtual Lab provides access to collections of speech and language data with a rich API to support data processing.
This page shows the different collections currently available on Alveo.
To access data from a collection, you must first register as an Alveo user (help me register).
You will then be able to view the licence terms for the collection and agree to them (about licence terms). In some cases specific permission is required.
To upload your own data, follow the instructions (help with uploading).
|ace||The ACE corpus was compiled to match with Australian data from 1986 the standard American and British corpora (Brown and LOB) from the 1960s. It includes 1 m. words of published text in 500 samples from 15 categories of nonfiction and fiction.|
|art||Australian Radio Talkback (ART) is a set of transcribed recordings of samples of national, regional and commercial Australian talkback radio from 2004 to 2006.. It includes 27 audio recordings and transcripts of talkback from ABC National Radio (NAT), ABC Radio broadcasts to eastern Australian (ABCE), ABC Radio broadcasts to southern and western Australia (ABCNE), as well as commercial stations broadcasting to eastern Australia (COME) and southern and western Australia (COMNE).|
|austalk||AusTalk is a large state-of-the-art database of spoken Australian English from all around the country. Collected from 2011-2013, almost a thousand adults of all ages from 15 different locations in all states & territories were being recorded, representing the regional and social diversity and linguistic variation of Australian English, including Australian Aboriginal English. Each speaker was recorded on three separate occasions to sample their voice in a range of scripted and spontaneous speech situations at various times.|
|austlit||AustLit provides full-text access to hundreds of examples of out of copyright poetry, fiction and criticism ranging from 1795 to the 1930s. The collection includes literature intended for popular audiences as well as literature intended for audiences concerned with literary quality or the establishment of a national canon. The bibliographical information associated with these records enables researchers to investigate the relationships between texts and particular publishers or to track the first publication of each text in newspapers, magazines or journals. This provides indirect evidence of the original audience for each text and the evolution of reception over time if the texts were subsequently republished in other contexts.|
|avozes||AVOZES is an audio-video (or auditory-visual) speech data corpus for Australian English. The AVOZES data corpus was designed and recorded with two major goals in mind. Firstly, a new framework for the design of comprehensive, well-structured, multiple-use AV speech data corpora was proposed and followed in the production of the AVOZES data corpus. Secondly, the first publicly available, comprehensive AV speech data corpus for Australian English (AuE) was produced. In addition, it is the first AV speech data corpus to use a stereo vision system.|
|braidedchannels||The Braided Channels research collection includes materials collected on Australia women, land and history in the Channel country. The collection is constructed from some 70 hours of oral history interviews with women from Australia's Channel Country, together with archival film, transcripts, photos and music.|
|cooee||Material to be included had to meet with a regional and a temporal criterion. The latter required texts to have been produced between 1788 and 1900 in order to become eligible for COOEE. It was mandatory for a text to have been written in Australia, New Zealand or Norfolk Island. But in a few cases, other localities were allowed. For example, if a person who was a native Australian or who had lived in Australia for a considerable time, wrote a shipboard diary or travelled in other countries.|
|gcsause||The Griffith Corpus of Spoken Australian English (GCSAusE) comprises a collection of transcribed and annotated recordings of spoken interaction amongst Australian speakers of English, as well as users of English in Australia more generally, collected by staff and students at Griffith University.|
|ice||The ICE-AUS corpus is a 1 m.word corpus of transcribed spoken and written Australian English from 1992-1995. Its internal structure with 500 samples (60% speech, 40%writing) matches that of other ICE corpora (associated with the International corpus of English.|
|irias||The IRIAS corpus (Italian Roots In Australian Soil) is a corpus of natural spontaneous speech collected following a sociophonetic approach in three languages (Dialect, Italian and English) from first and second generation Italo-Australian speakers originating from two specific regions in Italy (Veneto and Calabria). The aim is to extend and complement the more frequently investigated macro-structures of lexical, syntactic and morphological interactions among immigrants' languages and more common sociolinguistic investigations about immigrants' language attitudes. More specifically, the corpus was designed to investigate how certain phonetic features in the speech of two groups of speakers may be affected in the context of learning and using (and living in contact with) a new language in Australia.|
|jakartan_indonesian||Collection of Jakartan Indonesian conversations with transcripts.|
|llc||Transcribed Lecture Corpus|
|mava||The MAVA corpus (MARCS Auditory-Visual Australian recordings of IEEE sentences) is a collection of high quality audiovisual recordings of 205 phonetically balanced sentences from the IEEE sentence database, recorded by a native Australian English female talker. The audio channel is annotated at the word and phoneme level and the video channel is provided with a frame-by-frame lip contour tracking. The center of the lip region is used as a reference for deriving four video regions: full face, upper face, lower face and lips. All files are freely available for download under the Creative Commons BY-NC-SA licence.|
|mbep||96 recordings read by actors expressing 6 emotions. DOI: 10.13140/RG.2.1.4271.7600|
|mitcheldelbridge||The Mitchell and Delbridge database contains recordings of Australian English as spoken by 7736 students at 330 schools across Australia, mostly collected in 1960. The tapes were digitised and reissued as an online database by the University of Sydney in 1997-98.|
|monash||The Monash Corpus of Australian English (MCE) consists of a collection of recordings and transcriptions of interviews made in Melbourne c1997. The subjects of the interviews were adolescents from a variety of schools. The data were collected and transcribed by staff of the Linguistics Program, Monash University.|
|pixar||A selection of music excerpts from Disney-Pixar fantasy-adventure animations (Toy Story 3, A Bug's Life, Finding Nemo, Monsters Inc, Cars and Up). Includes at least three excerpts of each of the six target emotions (angry, calm, excited, happy, sad, scared). Nineteen excerpts are included ranging from 7 to 27 seconds in duration.|
|rirusyd||Doheon Lee, Densil Cabrera, Effect of listening level and background noise on the subjective decay rate of room impulse responses: Using time-varying loudness to model reverberance, Applied Acoustics, Volume 71, Issue 9, September 2010, Pages 801-811, ISSN 0003-682X, http://dx.doi.org/10.1016/j.apacoust.2010.04.005. (http://www.sciencedirect.com/science/article/pii/S0003682X10000861) Keywords: Room acoustics; Reverberation time; Reverberance; Impulse response; Loudness|
|sc-cw-children||Read speech collected in a recording studio from 8 children aged 6-12.|
|trove||The trove newspaper collection.|
New collections can be uploaded by researchers and shared with the community (help with uploading).