Prajwal R
Chief Scientist & Co-founder at University of Oxford
Bengaluru, Karnataka, India
Summary
🤩
Rockstar🎓
Top SchoolPrajwal R is a machine learning researcher and entrepreneur with 11 years of experience building state-of-the-art video understanding and generation systems, currently co-founder and Chief Scientist at sync. labs (YC W'24) in Bengaluru. His work sits at the intersection of vision and audio—he led core contributions to widely used projects like Wav2Lip and Lip2Wav, focusing on training, preprocessing pipelines, and evaluation that made lip-sync and lip-to-speech models practical and robust. As a doctoral student at the University of Oxford and a former Meta research intern, he has designed novel generative architectures (including text-to-music work at Meta) and published CVPR/ACM Multimedia papers on lip-speech synthesis and translation. Prajwal combines deep academic rigor with product-minded engineering, shipping usable inference options and hyperparameter tuning that improve real-world performance. He also has a proven track record across academia and industry—from medical imaging at NUS to computer vision and multimodal accessibility projects at IIIT Hyderabad—reflecting a broad applied-research portfolio. Notably, his contributions emphasize reproducibility and tooling (data preprocessing, scoring scripts, inference ergonomics), making cutting-edge models easier to adopt in production.
11 years of coding experience
3 years of employment as a software developer
Doctor of Philosophy - PhD, Artificial Intelligence, Doctor of Philosophy - PhD, Artificial Intelligence at University of OxfordBachelor of Technology (B.Tech.), Computer Science, Pre final year, Bachelor of Technology (B.Tech.), Computer Science, Pre final year at National Institute of Technology TiruchirappalliGrade 12th District topper, Science stream, 97.2 percent in the 12th CBSE exam, Grade 12th District topper, Science stream, 97.2 percent in the 12th CBSE exam at Suguna PipsMaster of Science - MS, Computer Science, Master of Science - MS, Computer Science at IIIT Hyderabad
Hindi, Kannada, Tamil, English
Stackoverflow
682reputation
57kreached
20answers
5questions
Github Skills (21)
pytorch10
python10
evaluation10
machine-learning10
speech-synthesis10
inference10
data-preprocessing10
eval10
trainings10
computer-vision10
modeling10
data-engineering9
audio-processing9
image-preprocessing9
tensorflow8
Programming languages (4)
Github contributions (5)
Rudrabha/Wav2Lip
Aug 2020 - Mar 2021
This repository contains the codes of "A Lip Sync Expert Is All You Need for Speech to Lip Generation In the Wild", published at ACM Multimedia 2020. For HD commercial model, please try out Sync Labs
Role in this project:
ML Engineer
ML EngineerContributions:3 reviews, 61 commits, 13 PRs in 6 months
Contributions summary:Prajwal primarily contributed to the training and inference code for a lip-sync generation model. Their work involved debugging and refining existing code in training scripts, fixing preprocessing steps, and adding inference options such as cropping and face box specifications. The user adjusted loss function weights and addressed issues related to audio indexing, showcasing a focus on model performance and practical usability.
lip-syncspeech
Rudrabha/Lip2Wav
May 2020 - Aug 2020
This is the repository containing codes for our CVPR, 2020 paper titled "Learning Individual Speaking Styles for Accurate Lip to Speech Synthesis"
Role in this project:ML Engineer
ML EngineerContributions:25 commits, 2 PRs, 12 pushes in 3 months
Contributions summary:Prajwal primarily contributed to the project by adding scripts for data preprocessing and downloading datasets for lip-to-speech synthesis. They developed a scoring script for evaluating the model's performance using metrics like PESQ and STOI. The commits also involved modifications to training scripts, including adding speaker-specific hyperparameters, and updating inference files, highlighting a focus on the model training and evaluation pipeline.
speech-synthesis
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.