An analysis of the 16th paper in the 2012 PRASA proceedings, covering its likely content, the state of pattern recognition in South Africa at the time, and the PDF as a historical digital artifact.
The Story Behind prasa2012_13.pdf: A Snapshot of Pattern Recognition Research in 2012
Exploring the context and content of paper #13 from the 2012 PRASA proceedings, covering language identification for under-resourced languages and the shift in research methodology.
How a 2011 PRASA Paper Tackled Language Identification for Under-Resourced South African Languages
Exploring the PRASA 2011 paper that used a UBM-MAP framework to identify Sepedi, isiXhosa, and Tshivenda with only 10 minutes of training data per language.
Decoding PRASA 2011 Paper #27: Speaker Diarization for Under-Resourced Languages
Explore the PRASA 2011 proceedings paper #27 on hybrid speaker diarization for under-resourced languages. A detailed historical analysis of mid-2000s speech technology.
Decoding the 24th Paper in PRASA 2011: A Snapshot of Under-Resourced Speech Research
Explore the 24th paper from PRASA 2011 proceedings, a study on acoustic feature fusion for language identification in code-switched under-resourced speech.
The 17th Paper at PRASA 2011: Subspace GMMs for IsiXhosa
An in-depth look at a 2011 PRASA proceedings PDF that documents early subspace GMM experiments for IsiXhosa speech recognition, capturing the state of the art before deep learning.
The Story Behind prasa2011_01.pdf: Pattern Recognition Research in 2011
A close look at the first paper from PRASA 2011 proceedings: its content, technical details, and what it reveals about speech technology for under-resourced languages in the early 2010s.
The PRASA 2009 Proceedings PDF: A Mid-2000s Pattern Recognition Artifact
Explore the PRASA 2009 proceedings PDF (prasa09_15.pdf) as a historical artifact of mid-2000s pattern recognition, speech technology, and conference production.
The Story Behind prasa08_29.pdf: A 2008 PRASA Language ID Paper
Explore the 2008 PRASA conference paper prasa08_29.pdf on language identification for low-resource South African languages, its technical details, and archival significance.
PRASA 2007 Proceedings: A Time Capsule of Mid-2000s Pattern Recognition Research
Explore the PRASA 2007 proceedings PDF—a snapshot of mid-2000s pattern recognition research, from speech processing to computer vision, before deep learning took over.