Research Scientist

Adobe Adobe · Enterprise · San Jose, CA

Research Scientist at Adobe Research focusing on generative world models for consistent world representations, long-horizon video and 3D generation, and interactive creative experiences. The role involves inventing algorithms, building prototypes, large-scale training, evaluation, and deployment, with potential integration into Adobe products. Requires a Ph.D. and significant research experience with a strong publication record.

What you'd actually do

  1. Pursue research in world models, including: controllable, interactive video and 3D generation; long-term temporal modeling and visual memory; spatially grounded scene representations
  2. Invent and prototype state-of-the-art algorithms in collaboration with Adobe researchers
  3. Publish in premier AI/ML, vision, and graphics venues and file patents
  4. Build functional prototypes, including large-scale training, evaluation, and deployment
  5. Work with product teams to integrate models into Adobe’s creative ecosystem

Skills

Required

  • Ph.D. in Computer Science or a related technical field, with a focus on AI/ML or equivalent industry experience
  • Minimum of four years of proven experience doing state-of-the-art research and translating that research into product impact
  • Expertise in one or more of the following areas: world models, video generation, 3D generation
  • A strong personal research track record including peer-reviewed publications in top-tier AI/ML, vision and graphics and/or contributions to open-source projects
  • Experience with large-scale generative/transformer models (data processing, distributed training, post-training and distillation, optimized inference, analysis and evaluation)
  • Exceptional Python and deep learning development skills and experience with one or more training or deployment environments (PyTorch, Tensorflow)
  • Strong verbal and written communication skills

Nice to have

  • contributions to open-source projects

What the JD emphasized

  • consistent world representations
  • long-horizon video and 3D generation
  • interactive, controllable creative experiences
  • world models
  • video generation
  • 3D generation
  • proven experience doing state-of-the-art research and translating that research into product impact
  • strong personal research track record including peer-reviewed publications in top-tier AI/ML, vision and graphics and/or contributions to open-source projects
  • large-scale training, evaluation, and deployment

Other signals

  • world models
  • video generation
  • 3D generation
  • long-horizon video and 3D generation
  • interactive, controllable creative experiences