How PRASA conference websites evolved from
tags to
layouts, revealing the early web’s struggle to describe physical spaces and what that tells historians of pattern recognition research.
Markup and Meaning: How HTML Shaped Early Document Analysis in Pattern Recognition
Explore the overlooked role of HTML in structuring documents for OCR, layout analysis, and early pattern recognition research in the 2000s.
When HTML Became the Default: Mid-2000s Pattern Recognition Conferences Go Digital
How mid-2000s pattern recognition conferences like PRASA moved from paper and PDF to HTML proceedings, boosting access for under-resourced language research.
The Quiet Backbone of PRASA: How Sponsorship Built South Africa’s Speech Research
Explore how national agencies, international programmes, and corporate sponsors funded under-resourced language research, data collection, and conference proceedings in the Pattern Recognition Association of South Africa (PRASA) from 2000 to 2010.
Registration HTML in the Mid-2000s: How Academic Conferences Adopted Web Forms
Explore the evolution of HTML registration forms in academic conferences during the mid-2000s, from simple tables to server-side processing.
The Hidden Workhorse of Mid-2000s Pattern Research: Registration Before Deep Learning
How registration—geometric, temporal, and non-rigid—held together computer vision, speech recognition, and medical imaging before deep learning took over, with a look at under-resourced language research from PRASA symposia.
PRASA 2002 Programme: A Window into Early 2000s Pattern Recognition
Analyzing the PRASA 2002 conference programme reveals the structure, priorities, and evolution of early 2000s pattern recognition research in South Africa.
What ‘Program HTML’ Meant in Mid-2000s Pattern Recognition Research
Explore how HTML was used to disseminate pattern recognition research in the mid-2000s, from conference proceedings to code examples, and why it was never a programming language.
Paper #48 at PRASA 2014: Language ID for Code-Switched South African Speech
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.