Pablo Vela
Computer Vision Engineer at Immersed [Techstars '17]
Austin, Texas, United States
Summary
🤩
Rockstar🎓
Top SchoolPablo Vela is a computer vision and robotics engineer with nine years of hands-on experience building and productionizing 3D perception systems, currently based in Austin, Texas. He has grounded academic training in mechanical engineering and self-driving systems with practical work shipping monocular and multiview 3D pose and room reconstruction pipelines at startups like Immersed and contributing to open-source NeRF and SDF tooling. Pablo’s contributions to popular projects such as nerfstudio—adding SDF mesh export, texturing, and NeusFacto—and sdfstudio—adapting data pipelines and adding monocular depth/normal support—showcase his ability to bridge research methods and engineering-grade implementations. He’s experienced with PyTorch, OpenCV, optimization techniques (e.g., Levenberg–Marquardt), and edge deployment using Google Mediapipe for real-time inference. Beyond models, he has a track record of practical systems work—from sensor integration and per-image intrinsics support to generating geometric cues that accelerate SDF optimization. Curious and interdisciplinary, Pablo pairs robotics-mechanical intuition with ML-first approaches to improve 3D reconstruction fidelity and speed.
9 years of coding experience
1 year of employment as a software developer
Self Driving Car Nanodegree, Deep Learning/Machine Learning/Computer Vision/Controls, Self Driving Car Nanodegree, Deep Learning/Machine Learning/Computer Vision/Controls at UdacityBachelor’s Degree, Mechanical Engineering, Bachelor’s Degree, Mechanical Engineering at University of Florida
English, Spanish, Spanish
Stackoverflow
31reputation
6kreached
0answers
1question
Github Skills (15)
3d-reconstruction10
computer-vision10
pytorch10
serf10
deep-learning10
python10
image-processing10
sdf10
data-processing10
3d10
machine-learning9
3d-graphics9
anaconda6
conda6
tensorflow6
Programming languages (7)
Github contributions (5)
nerfstudio-project/nerfstudio
Oct 2022 - Dec 2024
A collaboration friendly studio for NeRFs
Role in this project:
ML Engineer
ML EngineerContributions:17 reviews, 16 PRs, 48 pushes in 2 years 2 months
Contributions summary:Pablo primarily contributed to the development and improvement of the NeRFStudio project, specifically focusing on incorporating and refining SDF-based methods. Their work includes the addition of SDF mesh export and texturing capabilities, the implementation of the NeusFacto method, and integration of the Pixel Perfect SFM option. The user also addressed several bugs, implemented new loss functions, and made improvements in several of the project files.
pytorchphotogrammetrydeep-learningcomputer-vision3d-reconstruction
autonomousvision/sdfstudio
Dec 2022 - Jan 2023
A Unified Framework for Surface Reconstruction
Role in this project:ML Engineer
ML EngineerContributions:1 review, 7 commits, 7 PRs in 14 days
Contributions summary:Pablo primarily contributed to the development and refinement of data processing scripts for the `sdfstudio` framework. Their work involved adapting existing scripts from `nerfstudio` for use in `sdfstudio`, incorporating features like indoor scene handling and support for per-image intrinsics. The user also added support for generating monocular depth and normal maps using external tools like Omnidata, along with the sensor depth from Polycam.
pytorchreconstructionsurface-reconstructionsdfimplicit-neural-representation
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.