Senior Machine Learning Engineer, Video Quality Systems

As a Senior Machine Learning Engineer, you will tackle one of the most persistent challenges in video technology: reliably measuring perceived visual quality at scale. While human expert evaluation remains the gold standard for accuracy, it is resource-intensive and slow. Conversely, traditional automated metrics offer speed, but often fail to correlate meaningfully with human perception. You will be an expert in designing a hybrid evaluation framework. By leveraging large-scale outsourced subjective data, you will characterize the boundaries of existing automated metrics and inject domain and "world knowledge" to apply them only where they are statistically reliable. Ultimately, your goal will be to design and tune novel, explainable metrics. We are explicitly looking for an approach grounded in first principles of signal processing and human vision, rather than relying on opaque, "black-box" machine learning models that simply output a quality score. Your work will directly accelerate our core engineering efforts by providing developers with rapid, trustworthy, and actionable feedback. Subjective Testing & Analysis: Design, oversee, and analyze large-scale psycho-visual experiments to collect high-quality subjective video evaluation data. Metric Characterization: Evaluate existing objective Video Quality Assessment (VQA) metrics against human baselines to determine their correlation and operational limits. Context-Aware Evaluation: Develop methodologies to classify video content and apply "world knowledge," identifying exactly which automated metrics succeed or fail on specific types of content and artifacts. First-Principles Design: Design, tune, and validate new objective quality metrics based on the human visual system (HVS) and mathematical first principles, ensuring the resulting scores are highly explainable and actionable. Cross-Functional Collaboration: Partner with algorithmic development teams to integrate your evaluation frameworks into fast, automated feedback loops that guide the engineering process. MS in Machine Learning, Computer Science, Applied Mathematics, or a related discipline and minimum 10 years relevant industry experience. Demonstrated experience on Image/Video Quality Assessment (IQA/VQA), image processing, or computational vision. Track record in statistical analysis, correlation methodologies, and data modeling. Proficiency in algorithm architecture design and implementation. PhD in Machine Learning, Computer Science, Applied Mathematics, or a related discipline. Experience managing or scaling outsourced/crowdsourced subjective evaluation campaigns (e.g., using ITU-T standards). Track record of developing explainable, non-black-box algorithms for image or video analysis. Proven experience designing, conducting, and analyzing psycho-physical or psycho-visual experiments for subjective quality evaluation. Demonstrated knowledge of the human visual system (HVS), perceptual artifacts, and traditional signal processing, evidenced through publications, coursework, or applied project work. Working knowledge with modern video processing pipelines, compression standards, and enhancement algorithms. Strong publication record in relevant venues (e.g., VQEG, ICIP, HVEI, SPIE) or equivalent industry patents. Ability to translate complex perceptual phenomena into clear, actionable engineering requirements, as demonstrated through technical writing, presentations, or cross-functional collaboration.

For more details click Job Post.