Exploring a 2014 PRASA paper that tackled language identification in code-switched South African speech using acoustic and prosodic features.
Bottleneck Features for Language Identification: A Look at PRASA 2014 Paper #46
Explore a 2014 PRASA paper that used bottleneck neural network features for language identification across five South African languages, highlighting its handling of data scarcity and its place in speech technology history.
Inside PRASA 2014 Paper #38: Language ID for Under-Resourced Languages
Explore PRASA 2014 paper #38 on i-vector language identification for five under-resourced South African languages, with experimental details and historical context.
The Curious Case of the PRASA 2014 Proceedings Filename
A look at PRASA 2014’s oddly named PDF, its historical context, and what paper #20 tells us about pattern recognition research before deep learning.
Inside a Forgotten PDF: PRASA 2013 Paper 16 and African Language Speech Research
A deep dive into prasa2013_16.pdf reveals how South African researchers tackled language identification for under-resourced languages using GMMs and HMMs—before deep learning took over.
The 2013 PRASA Proceedings: Decoding ‘proceedings 2013 prasa2013 02 pdf’
Exploring the 2013 PRASA conference proceedings PDF: what the filename reveals about academic publishing, pattern recognition research, and the state of the field in 2013.
Inside prasa2012_33.pdf: How a 22-Page Paper Changed Language ID for South African Languages
A detailed look at a pivotal PRASA 2012 paper that used i-vectors and GMMs to improve language identification for five under-resourced South African languages, with full methodology, results, and lessons for modern researchers.
Inside the PRASA 2012 Proceedings: Paper #32 on Language Identification
Explore the PRASA 2012 proceedings paper #32 on language identification for under-resourced Bantu languages using subspace GMMs. Methodology, results, and historical context.
Decoding a 2012 PRASA PDF: What Eight Pages Say About Under-Resourced Language Research
A deep dive into the prasa2012_31.pdf from the 2012 PRASA symposium—exploring its filename, technical details, and what it reveals about language identification for Southern African languages.
Inside PRASA 2012’s 30th Paper: Code-Switching Language ID Before Deep Learning
A detailed look at the 30th paper from the PRASA 2012 proceedings, tackling code-switched language identification with SVMs, feature engineering, and a 78.3% accuracy baseline.