2026 Fall Intern, Computer Vision/AI

About the role#

Samsung Research America is seeking a Computer Vision/AI Intern to join the Vision Intelligence lab in Mountain View. This role focuses on advancing multimodal understanding and generation, with the opportunity to contribute to research that impacts future products. You will collaborate with researchers on projects involving image, video, 3D, text retrieval, question answering, reasoning, and generation.

What you'll do#

• Conduct research on multi-modal models.
• Write maintainable code to implement research ideas.
• Work with the team to discuss findings and refine methods.
• Execute experiments, document progress, and present insights.
• Prepare research findings for submission to major conferences or journals.

What you'll need#

• Currently pursuing an MS or PhD degree in Computer Science, Data Science, or Computer Engineering.
• Experience with Python, PyTorch, and Transformers.
• Previous publications in top-tier CV/ML/AI conferences or journals such as CVPR, ECCV, ICCV, ICML, NeurIPS, or AAAI.
• Experience in one or more of the following areas: VLMs (architecture design, pretraining, or post-training techniques like SFT, RLHF, GRPO), diffusion models for image/video/3D, efficient model development (knowledge distillation, mixture of experts), or Vision Language Action Models (VLA) for agentic systems.
• Strong oral and written communication skills.

Location & details#

• Location: Mountain View, California.
• Term: Fall 2026.
• Modality: On-site.
• Employment: Full-time internship.
• Compensation: This is a paid position with an hourly pay rate range of $48—$63 USD.