Research Scientist, Low Level Vision, Level 4

Location: Bellevue, India | Los Angeles, Canada | New York, United States | Palo Alto, Canada | San Francisco, Canada | Seattle, India

Department: Research

Employment Type: Full time

About Snap Inc

Snap Inc is a technology company. We believe the camera presents the greatest opportunity to improve the way people live and communicate. Snap contributes to human progress by empowering people to express themselves, live in the moment, learn about the world, and have fun together.

About the Role

Snap Inc is a camera company that believes the camera presents the greatest opportunity to improve the way people live and communicate. We empower people to express themselves, live in the moment, learn about the world, and have fun together through our core products: Snapchat, a visual messaging app; Lens Studio, an AR platform; and Spectacles, our AR glasses. Snap Research is the innovation engine driving paradigm shifts in communication and self-expression via cutting-edge camera technologies. Join our Computational Imaging Research Team in NYC as a Research Scientist, Low Level Vision, Level 4, where you'll enhance photo/video quality for Snapchat and simplify Lens creation for AR creators. Our 2025 focus includes image/video quality assessment, super-resolution, compression, editing, and animations like 3D avatars and video generation, leveraging computer vision, computational photography, 3D motion, and efficient VLMs. In this role, you'll propose groundbreaking low-level vision technologies to elevate AIGC videos and real-time camera experiences on Snapchat and Spectacles. You'll prototype efficient solutions for mobile devices, collaborate with product teams to reach hundreds of millions of users, mentor interns, and publish at top conferences. Thrive in Snap's creative culture by tackling challenges in on-device NN compression, video codecs, and AR innovations that redefine visual storytelling. We're seeking a visionary with a PhD in CS, ML, or related fields, deep expertise in deep learning/computer vision, and a passion for camera tech. Snap's 'Default Together' policy fosters dynamic in-person collaboration 4+ days/week in our vibrant NYC office. As an equal opportunity employer, we celebrate diverse backgrounds to build innovative AR products. Enjoy comprehensive benefits including medical coverage, parental leave, mental health support, and equity in our success.

What You'll Do

• Propose and develop innovative technologies to improve image and video quality, such as image/video super-resolution/restoration, image/video quality assessment, or advanced video codec
• Enhance the quality of AIGC-generated videos using low-level vision techniques, such as super-resolution and video frame interpolation
• Contribute to image editing, efficient vision-language model development, and related projects
• Partner with product teams to deliver technologies to hundreds of millions of Snapchatters
• Work on projects including on-device neural network (NN) compression or non-NN methods speedup
• Mentor research interns and guide junior team members
• Publish research at top academic conferences or contribute to video compression standards
• Experiment with new technologies impacting Snapchat, Lens Studio, and Spectacles
• Identify and define impactful research projects in low-level vision for camera-based AR experiences
• Prototype and implement solutions in Python and/or C++ for real-time mobile performance

Minimum Qualifications

• PhD in a related technical field such as computer science, statistics, mathematics, machine learning or equivalent years of experience
• Strong theoretical foundations of generative AI and practical experience training, tuning, and modifying generative models
• Strong technical foundation in deep learning and computer vision
• Proven ability to lead and mentor interns, PhD students, and junior researchers
• Ability to collaborate effectively with product teams
• Strong and clear communication skills
• Ability to stay at the forefront of low-level vision or computational photography advancements

Preferred Qualifications

• Hands-on experience with on-device NN compression, such as quantization, pruning, knowledge distillation, NAS, and/or other emerging techniques
• Successful record of publication in top-tier academic conferences and journals (e.g. CVPR, ICCV, ECCV, ICLR, NeurIPS, SIGGRAPH, TPAMI) or contributions to video compression standards
• Strong technical foundation in signal processing, statistics, machine learning, and computer vision
• Hands-on experience with state-of-the-art non-NN methods in image processing and their speedup on mobile devices, such as Metal for iOS, OpenGL or Vulkan for Android

Knowledge, Skills & Abilities

• Out-of-the-box thinking for innovative research ideas
• Strong prototyping, implementation, and programming in Python and/or C++
• Developing or accelerating NN methods including dataset preparation, architecture/loss design, and NN compression
• Non-NN image processing methods and mobile speedup techniques
• Fast learning and adaptation to new topics like image editing and efficient VLMs
• Staying at the cutting edge of low-level vision, computational photography, and image/video quality
• Defining impactful R&D projects in academic and product contexts
• Leading and mentoring interns and junior researchers
• Collaborating with cross-functional product teams
• Clear technical communication and presentation skills
• Deep learning model training, tuning, and deployment
• Computer vision for camera and AR applications
• Signal processing for imaging enhancements
• On-device optimization for mobile AR experiences

Our Benefits

• Paid parental leave
• Comprehensive medical coverage
• Emotional and mental health support programs
• Compensation packages that let you share in Snap’s long-term success
• Dynamic collaboration in a creative, innovative culture
• Opportunities to impact products used by hundreds of millions
• Professional growth through mentoring and top-tier research publication
• Office environment fostering in-person teamwork 4+ days per week

Compensation

$173,000-$259,000 annually

This position is eligible for equity in the form of RSUs.

"Default Together" Policy: At Snap Inc, we practice a "default together" approach and expect team members to work in an office 4+ days per week.

Snap is proud to be an equal opportunity employer.