I’m Sewade Olaolu Ogun, PhD. I build AI products at GetVocal AI in Paris.

I did my PhD in Computer Science at Inria Nancy and Vivoka, on generating diverse synthetic data for ASR training data augmentation. Before that, I completed the African Masters in Machine Intelligence (AMMI) at AIMS.

This blog is where I write about what I’m learning, mostly through the lens of multimodality.

Research interests

  • Speech language models
  • Generative text-to-speech systems
  • Automatic speech recognition systems
  • Dataset curation and augmentation
  • Large language models

CV

Download my CV (PDF) or see my LinkedIn profile. Papers and talks are on the Publications page.

News

  • Jun 2025Our paper on NaijaVoices was accepted at Interspeech 2025. It introduces over 1,800 hours of speech-text pairs for the three major languages in Nigeria.
  • Oct 2024I will give a practical talk to members of the AI Saturdays Lagos community on “Implementation and evaluation of a research paper”. Slides are available.
  • Oct 2024I successfully defended my PhD thesis, “Generating diverse synthetic dataset for ASR training data augmentation”. Thank you to everyone who supported me on this journey.
  • Jul 2024I will submit my thesis at the end of July, with the defense slated for October. I am on the lookout for new and interesting deep-learning positions in research and industry.
  • Jun 2024Two papers accepted at Interspeech 2024, on improving NER for accented speech and on TTS for African-accented English. I will present them with my colleagues in Kos, Greece.
  • Nov 2023Attended the Rencontres des Jeunes Chercheurs en Parole (RJCP 2023) workshop in Grenoble, where I presented one of my papers.
  • Aug 2023Attending Interspeech 2023 in Dublin to present my work on improving the diversity and naturalness of TTS systems.
  • Jun 2023Attended VivaTech, the annual technology conference on innovation and startups in Paris. Thanks to the Inria innovation lab for sponsoring my trip.
  • Jan 2023Presenting my work on curating high-quality data from crowdsourced datasets for TTS training at the Spoken Language Technology (SLT) workshop in Qatar.
  • Oct 2022Implemented some iterative phase retrieval algorithms that can be used as vocoders for TTS.
  • Apr 2022Attending the DeepLearn summer school in Guimarães, Portugal.
  • Aug 2021Starting my PhD at Inria Nancy and Vivoka in October, working on automatic speech recognition with limited data for embedded systems.
  • Jun 2021A group of 8 research students and I implemented the SAINT paper in a week. It was great working as a team to implement every detail of the paper.
  • May 2021Volunteering with TReND in Africa, a charity supporting scientific capacity building across Africa, to teach Python to beginners.
  • Apr 2021New blog post: Consider using UAR instead of Accuracy for Classification tasks.
  • Apr 2021Volunteering at the International Conference on Learning Representations (ICLR 2021).
  • Mar 2021Started a machine learning research internship with ubenwa.ai, working on predicting infant asphyxia with machine learning.
  • Feb 2021Took part in the qualifying round of Google Hash Code 2021; our team, Team OKAY, ranked in the top 25% of the leaderboard.
  • Jan 2021New blog post: How to create a speech dataset for ASR, TTS, and other speech tasks, aimed at the speech processing community.
  • Dec 2020Teaching assistant for the African Masters in Machine Intelligence programme at AIMS Senegal.
  • Sep 2020Finished the coursework for my master’s at AIMS AMMI, and looking for internships in NLP or speech recognition.
  • Aug 2020Taught a 2-day Python workshop for beginners with colleagues, with over 100 participants. Materials are in this GitHub repo.

Contact

Email: sogun [at] aimsammi [dot] org