Scan Video Algorithms Intern, Video Coding; Gaussian Splatting), Fall · risk Netflix · Los

Position: Scan Jul 30 Video Algorithms Intern, Video Coding (Gaussian Splatting), Fall 2026 Apply · 3/10 risk Netflix · Los Location: Northern Worth applying Netflix Los Gatos,California,United States of AmericaVideo Algorithms Intern, Video Coding (Gaussian Splatting), Fall 2026 Skills and tasksCustom CUDA Kernel Optimization for Gaussian RasterizationNeural Video Compression & Rate-Distortion TheoryPerceptual Quality Metrics for 3D Novel-View SynthesisSkill-to-task connections are not available yet.Tasks from the postingExplore GS model compression strategiesYou + AICharacterize trade-offs among GS model sizeYou + AIIdentify strategies to reduce training timeIdentify strategies to reduce training timeYour edgeDesign and implement a proof-of-conceptDesign and implement a proof-of-conceptYou + AIContribute to dataset needs for scenesContribute to dataset needs for scenesYour edgePrepare for this rolePractice an interview or plan your next steps with AI Coach.of this job is already automatable. 70% is classified as human or shared work.0 AI does it 3 You + AI 2 Your edge of 5 dutiesYour edge 3 thingsFormulating original algorithmic approaches to compress 3D Gaussian representations.Evaluating visual perceptual quality artifacts that standard automated metrics miss.The field growing 10+ yearsDemand for spatial computing and 3D/4D photorealistic rendering is accelerating across hardware vendors and streaming platforms. Efficient compression of novel-view synthesis models is essential for bandwidth-constrained consumer playback.03 Git Hub Copilot Writing PyTorch model wrappers and evaluation boilerplate for Gaussian Splatting

Qualifications

• Perceptual Quality Metrics for 3D Novel-View Synthesis

• Design and implement a proof-of-concept

• Design and implement a proof-of-concept

• Formulating original algorithmic approaches to compress 3D Gaussian representations

Responsibilities

• Custom CUDA Kernel Optimization for Gaussian Rasterization

• Neural Video Compression & Rate-Distortion Theory

• Identify strategies to reduce training time

• Identify strategies to reduce training time

• Contribute to dataset needs for scenes

• Contribute to dataset needs for scenes

• Evaluating visual perceptual quality artifacts that standard automated metrics miss

• Demand for spatial computing and 3D/4D photorealistic rendering is accelerating across hardware vendors and streaming platforms

• Efficient compression of novel-view synthesis models is essential for bandwidth-constrained consumer playback

• 03 Git Hub Copilot Writing PyTorch model wrappers and evaluation boilerplate for Gaussian Splatting