Oiwi Parker Jones
Also known as ʻŌiwi Parker Jones
Oiwi Parker Jones is an Oxford researcher who applies machine learning to neural data and speech decoding to develop brain-computer interfaces. He grew up speaking Hawaiian and completed an Oxford DPhil on Hawaiian computational phonology and morphology in 2010. With Brendan Shillingford, he coauthored a 2018 paper that presents a finite-state approach and a hybrid neural method for restoring missing characters in older Hawaiian writing. PNPL began at Oxford in 2023, and Parker Jones coauthored LibriBrain in 2025, MEG-XL in February 2026, and LibriBrain100 in August 2026. As of October 2026, he leads PNPL and serves as a Principal Investigator at the Oxford Robotics Institute and Hugh Price Fellow in Computer Science at Jesus College; his career also includes postdoctoral training in imaging neuroscience and applied AI and a Medicine lectureship at St Peter's College.
- Now at
- Oxford Robotics Institute and Jesus College, Oxford
Timeline
-
2010
A DPhil on Hawaiian
In 2010, he completes an Oxford DPhil on “A Computational Phonology and Morphology of Hawaiian”. He grows up speaking Hawaiian and focuses his doctoral research on natural language processing for low-resource languages.
-
2018
Restoring old Hawaiian characters
With Brendan Shillingford, he coauthors “Recovering Missing Characters in Old Hawaiian Writing” in 2018. The paper presents an end-to-end finite-state transducer and a hybrid method that combines an FST with a recurrent neural network language model for transliteration.
-
2023
Starting the Neural Processing Lab
The Parker Jones Neural Processing Lab is established in Oxford's Department of Engineering Science in 2023. It works across neuroscience, machine learning, language and robotics to develop brain-computer interfaces.
-
June 2025
LibriBrain for speech decoding
In June 2025, he coauthors LibriBrain, a speech-decoding dataset with over 50 hours of MEG recordings from one participant. The release includes standard data splits and baselines for speech detection, phoneme classification and word classification.
-
February 2026
Pretraining with long MEG contexts
In February 2026, he coauthors MEG-XL, which uses 2.5 minutes of MEG context per sample. The model matches supervised word-decoding performance with a fraction of training data.
-
August 2026
LibriBrain reaches 100 hours
He coauthors LibriBrain100 in August 2026, expanding the dataset to over 100 hours of MEG recordings of naturalistic speech. Its broad multi-subject set adds about 40 minutes of data from each of 32 participants.
-
As of October 2026
Leading PNPL at Oxford
As of October 2026, he is Group Lead of PNPL, Principal Investigator at the Oxford Robotics Institute and Hugh Price Fellow in Computer Science at Jesus College. His work applies machine learning to brain data and develops practical brain-computer interfaces, alongside speech and language technology for the Hawaiian-speaking community. His degrees include a BA and an Oxford DPhil, and his career includes postdoctoral training in imaging neuroscience at UCL and Oxford and applied AI at ORI, plus a Medicine lecturer role at St Peter's College.
Career
- NowUniversity of OxfordGroup Lead, Neural Processing Lab
- NowOxford Robotics InstitutePrincipal Investigator
- NowJesus College, OxfordHugh Price Fellow in Computer Science
- University College LondonPostdoctoral researcher in imaging neuroscience
- St Peter's College, OxfordLecturer in Medicine
Education
- 2010University of OxfordDPhil
Sources and credits
Photo: We have no published photo of Oiwi Parker Jones yet. Send us a free-licensed photo.
- Dissertations and Thesis - Hawaiʻi - Hawaiian Language Research - Research Guides at University of Hawaii at Manoa
- Advancing the future of speech prosthetics
- Recovering Missing Characters in Old Hawaiian Writing - ACL Anthology
- PNPL | Parker Jones Neural Processing Lab
- LibriBrain: Over 50 Hours of Within-Subject MEG to Improve Speech Decoding Methods at Scale
- MEG-XL: Data-Efficient Brain-to-Text via Long-Context Pre-Training
- LibriBrain100: One Hundred Hours of Broad and Deep MEG Data for Neural Speech Decoding at Scale
- Oxford Robotics Institute | People - Oiwi Parker Jones
- Oiwi Parker Jones - People - Engineering Science Department