Vighnesh Birodkar
Member Of Technical Staff at Odyssey
Montreal, Quebec, United States
Summary
🤩
Rockstar🎓
Top SchoolVighnesh Birodkar is a machine learning researcher and engineer with 13 years of experience specializing in video generative models, compression, real-time systems, and long-context generation. Currently a Member of Technical Staff in Montreal, he previously drove video model pre-training at Luma AI and built early video infrastructure and prototype models at Google that influenced production systems. His open-source contributions span TensorFlow, scikit-learn, and scikit-image—ranging from core object-detection improvements and ResNet architecture work to robust testing and classical image-processing algorithms. He favors rapid, simple experiments to uncover deep insights and brings a practical blend of research rigor and engineering discipline across both academic and production settings. An underappreciated strength is his long-term familiarity with video generation workflows dating back to 2017, giving him rare institutional memory of how models evolve from prototype to deployment.
13 years of coding experience
10 years of employment as a software developer
Master’s Degree Computer Science, Master’s Degree Computer Science at New York UniversityBachelor of Technology (B.Tech.) Computer Science, Bachelor of Technology (B.Tech.) Computer Science at National Institute of Technology, Goa
Hindi, Marathi, Konkani, English
Stackoverflow
357reputation
31kreached
6answers
0questions
Github Skills (30)
algorithm10
algorithms10
opencv10
python10
image-processing10
testing10
scikit10
machine-learning10
scikit-image10
data-structure10
numpy10
keras10
deep-learning10
tensorflow10
resnet10
Programming languages (6)
Github contributions (5)
scikit-image/scikit-image
Mar 2014 - Jun 2016
Image processing in Python
Role in this project:
Data Scientist
Data ScientistContributions:282 commits, 11 PRs, 310 comments in 2 years 4 months
Contributions summary:Vighnesh contributed significantly to the development of the `blob_dog` and `blob_log` functions within the scikit-image library, which is dedicated to image processing. Their work included implementing the Difference of Gaussian (DoG) and Laplacian of Gaussian (LoG) blob detection algorithms, as well as refactoring and adding test cases for these features. Furthermore, the user made improvements to the code's accuracy and documentation. The contributions demonstrate a focus on image processing techniques and code quality.
image-processingpythoncomputer-visionimage
tensorflow/models
Apr 2019 - Nov 2022
Models and examples built with TensorFlow
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:7 reviews, 79 commits, 7 PRs in 3 years 7 months
Contributions summary:Vighnesh refactored and added new architectures to the ResNet model, incorporating modifications such as numerical layer naming and adjustments to convolutional strides and classes. They also made internal changes to object detection model libraries and test files, and to the CenterNet meta-architecture. Further contributions involved adding options for summaries and checkpoints in a TF2 launch script and modifying the core of target assigners for object detection.
deep-learningtensorflow
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.