Research Scientist, Real-Time Video Generation
Core
Researching and developing real-time video generation models for interactive characters, focusing on consistency, expression, and efficient inference.
Role type
Senior IC research scientist (real-time video generation)
Builds
Real-time video generation models for interactive characters
Domain
Generative AI, Computer Vision, Real-time Systems
Deliverable
production ML models
Required skills
Video model training, fine-tuning, distillation, inference optimization, data curation, experimental design, latency analysis
Preferred skills
Streaming generation, speech-motion alignment, model evaluation
Technologies
Video generation frameworks, fine-tuning libraries, inference runtimes
Responsibilities
Develop models for responsive interaction with visual quality and consistency; explore character expression and motion coherence; adapt models via fine-tuning and post-training; optimize generation speed and efficiency; curate and develop training data; build evaluation metrics and comparisons; conduct research and share experiments.
Seniority
Senior, hands-on IC
