Vision Researcher - Multimodal Understanding & Generation in Foundation Models

Tencent • Bellevue, WA

70d • $149,000 - $279,800

About The Position

The role involves serving as a domain expert in computer vision and collaborating with researchers from other modalities to drive cutting-edge research in native multimodal foundation models. This includes novel architecture design and modeling for '2D + time' and '3D + time' scenarios. The position also requires exploring the training and design of large models for understanding and generating representations of the physical world, multimodal reasoning, and self-evolving continual learning. Staying up to date with the latest advancements in academia and industry is crucial, as is actively participating in international conferences and workshops and engaging with leading global research teams. Additionally, the role involves contributing impactful research outcomes to the open-source community or transferring technologies to internal product teams.

Requirements

• Master's or Ph.D. degree in Computer Science, Artificial Intelligence, Computer Vision, Machine Learning, or a related field.
• Proven multi-modal research experience in relevant areas, with familiarity with state-of-the-art technologies and a strong publication record in top-tier conferences or journals such as CVPR, ICCV, ECCV, NeurIPS, ICLR, or ICML.
• Proficiency with mainstream open-source tools and frameworks relevant to the field, and strong engineering skills to support research implementation; candidates with influential GitHub projects or contributions to high-impact open-source communities are preferred.
• Strong team spirit and ability to collaborate across disciplines, excellent communication skills, intellectual curiosity, and a goal-oriented, problem-solving mindset.

Responsibilities

• Serve as a domain expert in computer vision and collaborate with researchers from other modalities to drive cutting-edge research in native multimodal foundation models, including novel architecture design and modeling for '2D + time' and '3D + time' scenarios.
• Explore the training and design of large models for understanding and generating representations of the physical world, multimodal reasoning, and self-evolving continual learning.
• Stay up to date with the latest advancements in academia and industry; actively participate in international conferences and workshops, and engage with leading global research teams.
• Contribute impactful research outcomes to the open-source community or transfer technologies to internal product teams.

Benefits

• Sign on payment eligibility
• Relocation package eligibility
• Restricted stock units eligibility
• Medical benefits
• Dental benefits
• Vision benefits
• Life and disability benefits
• Participation in the Company's 401(k) plan
• 15 to 25 days of vacation per year (depending on tenure)
• Up to 13 days of holidays throughout the calendar year
• Up to 10 days of paid sick leave per year

What This Job Offers

• Career Level: Senior
• Industry: Broadcasting and Content Providers
• Education Level: Master's degree
• Number of Employees: 5,001-10,000 employees