MScAC Research Internship

Lab Summary:

AI Center-Toronto, Samsung Electronics, is located in Toronto, Canada. We are seeking outstanding interns to join our team focused on AI with applications to camera and imaging technology. Our center focuses on both short-term and long-term research targeting Samsung products with the potential to impact millions of users worldwide. We are dedicated to utilizing AI to advance the state-of-the-art in image enhancement and restoration, computational photography, and photo editing.  

Position Summary:

We are recruiting interns who share our passion for research and innovation in a collaborative environment. The ideal candidate should possess a strong background in AI, specifically focusing on areas such as computer vision, machine learning, photography, and mobile systems.

We are particularly interested in candidates with knowledge of image processing. Interest in photography is a plus. The candidate should consider themselves a strong hacker/systemsbuilder, and have expertise in Python and PyTorch, as well as strong motivation to learn or have knowledge of model optimization and deployment. Experience with Android is a strong plus.

Required Skills:

• Strong background in AI, specifically focusing on areas such as computer vision, machine learning, photography, and mobile systems
• Practical deep learning experience
• Solid foundation in at least two programming languages
• Self-motivated, independent, and able to work well in a fast-paced collaborative environment
• Passionate about learning
• Strong communication skills (written and verbal).

Special Attributes:

• Interest and experience in photography
• Android development experience.

Multiple internship positions are available. More details on potential projects are listed below.

Project Details:

Camera image quality is a high priority area for Samsung. Several projects at AIC-Toronto are focused on improving image quality and appearance (also known as image restoration and enhancement) in challenging scenarios, including low-light, nighttime photography, and digital zoom. The primary hardware responsible for in-camera processing is known as an image signal processor (ISP). Our team at AIC-Toronto is focused on all aspects of ISP performance, from full software-based ISP emulation, to data augmentation and synthesis for training deep-learning ISP components. Moreover, the restored image has to be photo-finished and enhanced in a perceptually pleasing manner, making the image look crisp and vivid, adapting to different users' preferences. This is a great opportunity to learn how your smartphone camera really works and contribute to improving imaging for millions of users.

The aforementioned image restoration tasks require solving hard inverse problems, recovering a high-quality high-resolution image from one or more low-quality low-resolution images. This requires not only accurate modeling of the camera pipeline and data, but also modern deep learning tools for image synthesis and perceptual quality improvement. We are looking to apply modern techniques, such as vision transformers, diffusion or flow-matching models, as well as multi-modal LLMs for image restoration and enhancement tasks. Finally, we need to deploy these models on mobile phones with limited computational resources. 

Work Environment:

The candidate will have a chance to work with other top researchers and engineers, across the Samsung AI Centers around the world. The candidate will have the opportunity to publish at top-tier conferences and journals. This position is a great fit for engineers who want to be at the cutting edge of AI research and at the same time have direct real-world impact.