About the role#
The Seed Vision Team develops foundational models for visual generation. This internship focuses on research and application development to solve fundamental computer vision challenges in generative AI. You will work on foundational models for images and videos, aiming to improve interactivity, controllability, and pattern recognition in generated content.
What you'll need#
- Currently pursuing a PhD in Computer Science, Computer Vision, Machine Learning, or a related field.
- Research experience in generative modeling, specifically regarding video, motion, or temporal sequences.
- First-author publications in major conferences such as CVPR, ICCV, ECCV, NeurIPS, ICLR, or ICML.
- Proficiency in deep learning frameworks and experience working with large-scale video datasets.
- Experience with diffusion or transformer-based video models or long-context sequence generation.
- Familiarity with long-form video datasets and an understanding of perceptual metrics or user-study-based video evaluation.
What you'll do#
- Develop scalable architectures for long-range video generation that maintain consistent motion, identity, and layout.
- Explore hierarchical or recurrent latent structures to support generation across long temporal spans.
- Address technical challenges such as temporal drift, motion collapse, and the retention of high-frequency details.
- Investigate autoregressive or chunked generation strategies to balance output quality with memory constraints.
- Design evaluation protocols for long video quality, focusing on realism, consistency, and semantic continuity.
Location & details#
- This role is located in San Jose, California.
- This is a full-time, on-site internship.
- Applications are reviewed on a rolling basis.
- Candidates must be authorized to work in the United States. Sponsorship is not provided for this position.


