AI Researcher – Image Quality Metrics

AI-Based Image Quality Metrics Researcher 

Job Purpose

•  Be responsible for researching Image and Video Quality Metrics solutions targeted for real-time mobile game rendering.
• Integrate AI algorithms targeted for enhancing Computer Graphics pipeline both in efficiency and quality.

Key Responsibilities

• Research and develop new Image and Video Quality Metrics for different types of game graphics distortions.
• Researching, designing and developing comparison studies between Full-Reference vs Non-Reference vs Partial-Reference metric methods.
• Tackling technical challenges and reshaping AI research solutions to become product compatible solutions.
• Identify new AI technologies and plan for future projects and product solutions.
• Contribute to project requirements and understand AI product deployment tradeoffs.
• Establishing and creating datasets that meet product deployment scenarios. 
• Assess our algorithm solution robustness under different deployment scenarios.
• Analyse and improve efficiency and performance of solutions for either cloud or on-device deployment.
• Documenting and reporting progress to your team/senior management and to cross-location/functional teams.
• Collaborate with cross-functional/location managers, researchers and engineers.

Person Specification

Required

• Master/PhD degree in Machine Learning/Computer science/computer vision or related technical domain.
• 4+ years of industry experience working on projects in: computer vision, image and video quality metrics, deep learning, machine learning, graphics.
• Expertise in AI, Machine Learning and Deep Learning
• Experience developing systems for manipulating image/video and multi-modality content.
• Experience in collecting, cleaning and creating datasets for AI model development.
• Minimum 5+ years’ experience in at least one of the deep learning frameworks (e.g., Tensorflow, Caffe2, Pytorch, MxNet, Torch, etc.). 
• Record of publications in top-tier conferences: CVPR, ECCV/ICCV, SIGGRAPH, BMVC, NeurIPS, ICML, ICLR.