Video Algorithms Intern, Video Coding (Gaussian Splatting), Fall 2026
Core
Investigate Gaussian Splatting (GS) as a future streaming format and build a practical system for photorealistic novel-view synthesis on consumer devices.
Role type
PhD-level Video Algorithms Intern (3D/4D Scene Reconstruction)
Builds
GS-based rendering systems and compression strategies for streaming
Domain
Video streaming, 3D/4D computer vision, neural graphics
Deliverable
production ML models | product features
Required skills
Python programming, 3D/4D scene reconstruction, novel-view synthesis, Gaussian Splatting, NeRF, differentiable rendering, neural graphics, 3D computer vision, machine learning, deep learning, model training and evaluation
Preferred skills
real-time rendering, GPU programming (CUDA, WebGL), video compression, codec standards (HEVC, AV1), open-source multimedia projects, large-scale distributed systems, cloud computing
Technologies
Python, CUDA, WebGL, HEVC, AV1
Responsibilities
Explore GS model compression strategies using open datasets, characterize trade-offs among GS model size, training time, and rendered quality, identify and experiment with strategies to reduce training/encoding time, design and implement a proof-of-concept showcasing GS-based rendering
Seniority
PhD candidate (expected graduation June 2027+)
