Staff ML Engineer - BEV & Multi-Modal Perception
Is this job for you?
Build my CV Build your CV and see how well you match this role — and every other one.
Track your applications on mobile The free Whileresume app, on iPhone and Android.
The role
Lead BEV-based perception model development and define the technical roadmap across detection, segmentation, road topology, and scene understanding.
Design advanced multi-modal architectures that fuse camera, LiDAR, radar, and HD maps into unified 3D representations.
Develop foundational perception models using BEV transformers, voxel-based encoders, or implicit scene representations.
Own large-scale training workflows, including data sampling strategies, augmentation pipelines, distributed training, and hyperparameter optimization.
Improve robustness and generalization for long-tail conditions, and establish evaluation frameworks for geometric accuracy, temporal stability, and cross-domain transfer.
Collaborate with calibration, mapping, and fusion teams; mentor ML engineers and advance state-of-the-art through self-supervised learning and large-scale pretraining.
Design advanced multi-modal architectures that fuse camera, LiDAR, radar, and HD maps into unified 3D representations.
Develop foundational perception models using BEV transformers, voxel-based encoders, or implicit scene representations.
Own large-scale training workflows, including data sampling strategies, augmentation pipelines, distributed training, and hyperparameter optimization.
Improve robustness and generalization for long-tail conditions, and establish evaluation frameworks for geometric accuracy, temporal stability, and cross-domain transfer.
Collaborate with calibration, mapping, and fusion teams; mentor ML engineers and advance state-of-the-art through self-supervised learning and large-scale pretraining.
See the full job post
Responsibilities, requirements, skills and benefits — create your free account.
or
Already have an account?
Log inYou might also like these jobs
No closely matching jobs yet — here are the most recent ones.