Allan Zelener
Research Engineer at Zoox Inc.
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolAllan Zelener is a Research Engineer in New York with a Ph.D. in Computer Science and a decade of experience building computer vision and machine learning systems for 3D scene understanding and object recognition. He develops algorithms that turn 3D camera data into actionable scene-level reasoning and has applied those skills at Zoox after internships at Google and Twitter focused on object detection and attention-based image captioning. Allan is the maintainer/contributor of YAD2K, a Keras port of the YOLO object detector, where he implemented loss functions, architecture tweaks, and modernized the codebase to Keras 2 to improve training reliability. He also taught data structures and algorithms as an adjunct lecturer, reflecting a strong foundation in theory alongside applied research. Known for bridging research and engineering, he prefers projects that move novel 3D vision ideas from prototype to robust, production-ready implementations.
10 years of coding experience
1 year of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at City University of New York Graduate CenterBaruch College, City University of New York (CUNY)Bachelor's Degree, Computer Science, Summa cum laude, Bachelor's Degree, Computer Science, Summa cum laude at City University of New York-Hunter College
Russian
Stackoverflow
184reputation
99kreached
4answers
0questions
Github Skills (11)
object-detection10
keras10
computer-vision10
machine-learning10
tensorflow10
python10
numpy9
deep-learning9
qt-core6
anaconda6
transformation6
Programming languages (3)
Github contributions (5)
allanzelener/YAD2K
Feb 2017 - Jul 2017
YAD2K: Yet Another Darknet 2 Keras
Role in this project:
ML Engineer
ML EngineerContributions:47 commits, 5 PRs, 36 pushes in 4 months
Contributions summary:Allan focused on implementing and refining the YOLO (You Only Look Once) object detection model within the Keras framework. Their work included defining the loss function, making adjustments to the model's architecture, and correcting preprocessing steps for improved performance. The user also updated the project to the Keras 2 API and worked on a sample training script to test the model.
deep-learningyet-anotherdarknettensorflowkeras
allanzelener/torch-autograd
Jul 2016 - Sep 2016
Contributions:12 pushes, 8 branches in 1 month
pytorchautogradtorchmachine-learning
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.