Machine Learning Engineer – Neural Network Optimization, Autopilot AI

Key Skills

Java Programming, Python Programming, C++, SQL, Machine learning techniques, Deep Learning, Jupyter Notebook, NumPy, SciPy, Scikit-learn, CUDA/GPU programming, matplotlib, Linux Operating system, Bash scripting, Computer Vision (CV)

Job Description

You will be part of a team of computer vision experts that research, design, implement, optimize, and deploy neural networks and computer vision algorithms that advance the state of the art in perception for autonomous driving.

What You’ll Do

• Train neural networks for autonomous driving tasks, such as object detection, pose estimation, depth estimation, keypoint detection & tracking, SLAM, etc.
• Participate in the complete computer vision task life cycle: develop a new task or neural architecture, optimize it for execution on our in-vehicle compute platform, deploy it to a development build, work with peer engineering teams to integrate it into our planning and controls stack, and validate its safety with our QA team.
• Stay abreast of advances in machine learning, particularly in the areas of object detection, object tracking, depth estimation, SLAM, neural radiance fields (NeRF).

What You’ll Bring

• Strong experience with Python, C++, and software engineering best practices.
• An “under the hood” knowledge of deep learning: layer details, loss functions, optimization, etc.
• Fluency with common neural network architectures for computer vision.
• Expertise in one of these areas: object detection, object tracking, pose estimation, depth estimation, multi-view stereo architectures, neural radiance fields (NeRF), video networks, keypoint detection & tracking, SLAM.
• Experience with PyTorch, or another major deep learning framework such as TensorFlow.
• Experience with data manipulation tools including jupyter notebooks, numpy, scipy, matplotlib, scikit-learn, Python & Bash scripting, Linux environment.
• (bonus) Experience with SIMD/SIMT parallelism, GPU programming, multithreaded programming.
• (bonus) Comfortable working in a shared cluster environment (e.g. Slurm, network filesystems).
• (bonus) Record of driving research projects from start to end, including conception, experimentation, iteration, and integration into a working product.