Computer Vision Engineer
About the position
We are looking for a Middle Data Science / Computer Vision Engineer to join a team developing autonomous UAV capabilities for challenging operating environments, including scenarios with limited or unavailable GNSS and operator connectivity.
The role is focused on Computer Vision and Machine Learning for real-world autonomous systems. You will work on algorithms for optical navigation, object detection and tracking, and other perception tasks, with a strong emphasis on deployment to onboard computing platforms and validation in simulation, on the bench, and during flight tests.
We are looking for an engineer who is comfortable working beyond the model itself - analyzing system behavior, investigating test results, identifying root causes, and turning prototypes into reliable solutions that can operate on real UAV platforms.
Must have skills:
- Middle-level experience: approximately 2+ years of practical Data Science / Machine Learning experience with Computer Vision as the primary specialization
- Strong Python skills
- Hands-on experience with PyTorch and/or TensorFlow
- Hands-on OpenCV experience
- Good understanding of modern CNN- and Transformer-based CV architectures
- Practical experience in several of the following areas: object detection, object tracking, classification, segmentation, feature extraction/matching, image registration, motion estimation
- Understanding of Computer Vision geometry: camera models, coordinate systems, camera calibration, transformations, and projective geometry
- Good foundation in linear algebra, probability, and statistics
- Experience with Linux, Git, and sound software engineering practices
- Ability to work with imperfect real-world data collected from field systems and flight tests
- At least an Upper-Intermediate level of English (spoken and written)
Your tasks will include:
- Develop and improve Computer Vision algorithms for UAV applications
- Develop object detection, classification, and tracking algorithms
- Work on optical navigation, Visual Odometry / Visual-Inertial Odometry, feature matching, image registration, and visual localization
- Train, fine-tune, and evaluate models using proprietary datasets
- Build and optimize real-time CV pipelines
- Optimize inference for edge platforms with limited compute resources
- Integrate DS/CV components with navigation, flight controllers, telemetry, and guidance logic
- Participate in simulation, bench, and flight testing
- Analyze failed tests and algorithm degradation, and improve robustness
We offer:
- Delivering high-end software projects that address real-world problems
- Surrounding experts who are ready to move forward professionally
- Professional growth plan and team leader support
- Taking ownership of R&D and socially significant projects
- Participation in worldwide tech conferences and competitions
- Taking part in regular educational activities
- Being a part of a multicultural company with a fun and lighthearted atmosphere
- Working from anywhere with flexible working hours
- Paid vacation and sick leave days