About the Role
The role involves creating and refining neural models that reconstruct 3D environments from sensor inputs to improve the accuracy and reliability of autonomous driving systems.
Responsibilities
- Design and implement neural networks for 3D scene reconstruction
- Develop algorithms that process LiDAR, radar, and camera data
- Optimize model performance for real-time inference
- Collaborate with engineering teams to integrate models into full-stack systems
- Evaluate reconstruction accuracy under diverse driving conditions
- Publish findings in top-tier research venues
- Stay current with advancements in computer vision and deep learning
- Improve training pipelines using large-scale sensor datasets
- Contribute to data curation and annotation strategies
- Troubleshoot model failures in complex urban scenarios
- Work closely with simulation teams to validate system behavior
- Refine loss functions to enhance geometric fidelity
- Support deployment of models on embedded hardware
- Analyze edge cases involving occlusions and dynamic objects
- Enhance robustness to adverse weather and lighting
- Drive research initiatives from concept to production
- Mentor junior researchers and engineers
- Ensure models generalize across geographies and vehicle types
- Collaborate on cross-functional safety validation efforts
- Maintain high standards for code quality and reproducibility
Nice to Have
- Postdoctoral or industry research experience
- Work involving real-world autonomous systems
- Contributions to neural radiance fields (NeRF)
- Experience with temporal consistency modeling
- Involvement in large-scale autonomous driving projects
- Knowledge of occupancy grid networks
- Background in geometric deep learning
- Familiarity with safety-critical software standards
Compensation
Competitive salary and equity package
Work Arrangement
Hybrid or remote options available
Team
Interdisciplinary team focused on autonomous vehicle technology
What We're Building
- Developing a next-generation perception system that uses neural reconstruction to model the 3D world with high fidelity
- Creating scalable solutions for interpreting complex urban environments in real time
Culture & Values
- Emphasis on technical excellence and scientific rigor
- Fast iteration cycles between research and deployment
- Open collaboration across teams with diverse expertise
Available for qualified candidates
