Topic – Transcription, text recognition & cultural heritage computing
Dr Steven Mills is a Senior Lecturer in the Department of Computer Science. His research is in computer vision, using computers to extract useful information from images and videos. He has a particular interest in cultural and heritage applications, including collaborations with archaeologists, archivists, and artists. He will present the results of preliminary work using deep neural networks to recognise letters and words in handwritten documents from the Marsden Online Archive. He will also attempt to explain what deep neural networks are, apart from “very mysterious”.
Lynn Benson is the Researcher Services Manager for Hocken Collections. She will explore some international initiatives and developments that are possible paths for the Library to follow in our goals to improve delivery of our digitised and born-digital collections.
Reading – Here’s How Google Deep Dream Generates Those Trippy Images | Madison Margolin
An introduction to the images produced by Google’s Deep Dream computer vision platform with an excellent video explanation by Dr Mike Pound.
You can even try generating your own images with one of the online Deep Dream generators: https://deepdreamgenerator.com/
Projects – READ, Visualize the Public Domain, Venice Time Machine, Arabic Scientific Manuscripts, Gravitron
- Recognition and Enrichment of Archival Documents (READ) – an eInfrastructure project by the European Commission to develop cutting-edge technologies and make archival material more accessible.
- Visualize the Public Domain | New York Public Library – an experiment by NYPL Labs to help patrons understand and explore the more than 180,000 images released by NYPL into the Public Domain.
- Venice Time Machine – History and Big Data – a project to digitise the 80 kilometres of Venetian state archives to allow researchers to reconstruct the city of the past.
- Arabic Scientific Manuscripts | British Library – a project to train software to read historical Arabic manuscripts.
- Gravitron – the work and thoughts of Geoff Hinchcliffe, senior Design lecturer at ANU, about design, data, computation and interface aesthetics.
WHEN: 12pm – 1pm, Friday 15th Feburary 2019
WHERE: Digital Humanities Hub, Room 1W3, First Floor, Arts Building
WHO: Anyone in the University community – there’s no advance registration required, but we always appreciate knowing in advance if you are planning to come along!
CONTACT: Alexander Ritchie alexander.ritchie@otago.ac.nz
One thought on “[Open Hours] Transcription, text recognition & cultural heritage computing”