Machine Learning Engineer, Computer Vision - McLean, VA, United States - MetroMSR - Staffing Resources

Description

About The Team

Our group specializes in developing a biometric iris recognition system scalable for a vast user base. With the latest machine learning techniques integrated into specialized hardware, we ensure high-resolution image capture, accurate identification, and fraud deterrence. This is achieved while maximizing user convenience. Our emphasis on safeguarding user data privacy helps enhance system efficiency and diminish model bias. Role Overview This position is pivotal in ensuring our biometric device, termed 'the Orb', offers a superior user experience by capturing exceptional biometric data. The unique imaging system, which adjusts its viewpoint for optimal iris capture, is crucial for acquiring the needed quality of data. Your primary responsibility will involve steering the software directing the imaging system, encompassing the evolution of computer vision models, their amalgamation into the hardware, and sourcing additional data for enhancing our model's capabilities. Key Responsibilities

• Collaborate with teams overseeing hardware, firmware, and data collection to curate workflows for the required training data.
• Develop bespoke data labeling services to elevate the caliber of our training data.
• Construct and train neural architectures to tackle challenges such as localization or semantic differentiation.
• Devise strategies to assess engine performance, pinpoint shortcomings, and ameliorate them, perhaps through data augmentation.
• Certify the uniform performance of the biometric system across diverse user demographics.
• Develop visualization tools and oversight systems for monitoring image quality, data anomalies, and the robustness against rare cases.
• Conduct experiments and delve into research on topics such as pupil contraction response times.

About The Ideal Candidate

• Professional experience or a PhD centered on computer vision applications leveraging deep learning, ideally with a history of deployed projects.
• Proficient knowledge of image processing techniques such as edge discernment, alignment, image quality metrics, noise reduction, and image merging.
• A keen eye for data and label precision.
• Mastery over Python and deep learning toolkits like Pytorch or Tensorflow.
• Familiarity with the current best practices in deep learning applied to computer vision.
• Capable of comprehending scientific literature, replicating findings, and adapting methodologies across contexts.
• Bonus: Acquaintance with Rust and experience working with MongoDB, PostgreSQL, and AWS platforms.

#J-18808-Ljbffr