Research Intern Multimodal Foundation Model for Vision

Research Intern – Multimodal Foundation Model for Vision 

Sony AI is seeking research interns to join us. Our team mainly focuses on fundamental and applied research, with a focus on building next-generation foundation models for vision in a responsible manner. The role of a research intern is to develop efficient and effective methodologies and prototype solutions. You will work with a productive team of world-class scientists and engineers to tackle the most challenging problems in foundation models and generative AI, including low-cost yet powerful vision foundation models (VFM), vision-language models (VLM), unified models, automatic model compression, optimization and deployement on cloud and edge. You will see your ideas not only published in papers, but also improve the experience of billions of customers. 

Roles and Responsibilities 

• Conduct fundamental and innovative development in low-cost yet powerful vision-language models (VLM), unified models, automatic model compression, optimization and deployement on cloud and edge. 

• Design or implement state-of-the-art techs on model compression, inference speedup, deployement on harwares, tool automation. 

• PoC for various vision+text, generation relevant tasks (VQA, captioning, understanding, etc) and hardwares. 

• Contribute to library and tool development to support business; or Publish influential research in top-tier conferences and journals. 

Required Qualifications and Skills 

• Currently has, or is in the process of obtaining, a master/PhD degree in computer science or related field. 

• Be very self-motivated and capable of proposing and implementing innovative ideas. 

• Solid presentation and communication skills to internal and external audiences. 

• Publications or expertise in compact foundation model development and deployment. Influential open-source projects or paper publication at top conferences, e.g., CVPR, ICCV, ECCV, NeurIPS, ICML, ACL, etc. 

• Better to have front-end development experience. 

• Solid coding skills in Python, Pytorch, etc.

​ 

Working Location 

Location flexible (Tokyo, Europe , US) 

The target hourly rate for this internship is $50.00 per hour. The individual will be paid hourly and eligible for overtime.   

#LI-AS1