Tech Stack

Description

You will be directly responsible for training, evaluation, and deployment of ML-based scene understanding, behavior prediction, and planning models for self-driving vehicles. Design experiments, improve evaluation metrics, and release major model updates while staying current with industry trends.

Requirements

• In-depth understanding of ML and deep learning algorithms
• Experience designing and training neural networks for motion planning, object detection, sensor fusion, motion prediction, or multi-object tracking
• Strong track record of tackling less-well-defined problems efficiently
• Advanced knowledge of software engineering principles (design, version control, build processes, code reviews, testing)
• Experience with PyTorch or other Python-based deep learning frameworks

Responsibilities

• Design and execute high-value experiments collaboratively
• Prototype and implement continuously improving evaluation metrics
• Timely release of major model updates with off-line and on-road evaluation analysis
• Stay up to date with latest trends and propose new architectures based on literature
• Maintain high-quality training and evaluation code-base including dataset generation