Algorithm Group Manager - Close Range 3D Perception
We are the Close Range 3D Perception group at Mobileye, made up of multiple teams of ML engineers and researchers.
We own HawkEye, Mobileye's next-generation model for close-range perception: an end-to-end 3D perception model that understands the area close to the autonomous vehicle. It fuses multiple surround cameras into a single 3D representation of the scene, and from it outputs the 3D structure of the environment, parking spots, road layout and markings, and road and camera conditions.
HawkEye is a transformer-based model built on a BEVFormer-based architecture. It is the perception layer behind automated parking and low-speed maneuvering, so its accuracy and robustness directly shape the safety of the product.
HawkEye is in its early stages, and the work is dynamic: priorities and requirements change as the product takes shape. In parallel, the group is already developing the future generation of HawkEye, based on a different approach.
We are looking for a Director to lead the group: own HawkEye's technical direction and roadmap, lead the team managers and engineers, and deliver to production programs.
What will your job look like:
- Lead multiple teams of ML engineers and researchers, through team managers and technical leads.
- Own the technical direction and roadmap of HawkEye and its future generation.
- Drive models from research through POC into production on Mobileye's EyeQ platforms, meeting accuracy, robustness and runtime requirements.
- Make the key architecture, data and evaluation decisions, and stay close to the algorithmic details.
- Work with partner groups (backbone, data, validation, system and product teams) and represent the group to customers and management.
- Hire, grow and retain a strong team.
All you need is:
- M.Sc. in Computer Science, Electrical Engineering, Mathematics, or a related field (PhD - an advantage).
- 10+ years of experience developing deep learning models for computer vision.
- Proven experience taking vision models from research into production.
- Experience managing engineers or teams.
- Strong technical judgment and the ability to set direction in a fast-moving field.
Nice to have:
- Proven experience in 3D perception or a closely related field (e.g., BEV / multi-camera perception, occupancy, 3D detection, depth estimation, 3D reconstruction).
- Experience with automotive perception, autonomous driving, parking or robotics.
- Experience with sparse or query-based 3D perception architectures, transformers, and multi-camera fusion.
- Experience deploying models on embedded or edge hardware.
- Publications at leading conferences (e.g., CVPR, ECCV, ICCV, NeurIPS, ICML, ICLR).
