Research Intern – MSR Asia Visual Computing Group

Key Skills

Java Programming, Python Programming, SQL, Machine learning techniques, Data Modeling, R Programming, Algorithms, CNNs, Mathematics, 3D Graphics

Job Description

The Visual Computing Group at Microsoft Research Asia consists of an elite team of researchers whose expertise spans the entire spectrum of research topics in computer vision, from mathematical theory to practical applications, from physical systems to software development, from image understanding to image generation, from 2D vision to 3D vision, and from pure vision processing to visual-language modeling. Representative research works from the group include ResNet, Faster R-CNN, DCN, HRNet, Face X-Ray, and Swin Transformer, which have made strong values to both the research community and internal business.

Roles & Responsibilities

We are looking for several interns in the Visual Computing Group of Microsoft Research Asia. Each research intern will work directly with one or more experienced researchers (i.e., the research mentors) in our group to conduct cutting-edge research in computer vision. Results produced from the research projects will not only lead to research publications in top conferences/journals, but also have the potential to contribute to various software products inside Microsoft.

As a research intern, your will work closely with your research mentor(s) to:

• identify the research problems and define your research project;
• study the literature, conceive the solutions, implement the algorithms in high quality research code, and conduct comprehensive experiments for validation;
• write technical reports, which will eventually transform to paper submissions to top conferences/journals in computer vision/machine learning/computer graphics;
• participate in research reading/discussion groups and seminars, and contribute to the various research activities in the group;
• enjoy the various trainings/benefits we provide to our research interns!!

Main Research Directions and Projects

• Designing of new vision backbones, especially the Transformer based architectures
• Self-supervised visual representation pre-training
• Visual-linguistic pretraining and applications
• Generative models, including Generative Adversarial Network and Denoising Diffusion Model and their applications, e.g., image enhancement and text-to-image generation
• 3D vision and content creation, including single-view / multi-view reconstruction, neural rendering, 3D representation learning, 3D controllable image generation, virtual and digital avatars
• Your research topic is not limited to the above directions; all research problems related to computer vision are welcomed if it is of sufficient interest.

Required Qualifications

• Ph.D./M.S./Bachelor degree (or actively pursuing) in Computer Science, Electrical Engineering, Automation, Applied Mathematics, or a related field
• Demonstrated academic credentials
• Relevant past research experience is a plus
• Solid mathematical background
• Strong programming skills
• Strong communication skills
• Self-motivated in conducting research
• Good spoken and written English skills
• Attention to details
• Open-minded with collaborative spirit