The Perception Vulnerable Road Users (VRU) Understanding team focuses on identifying, understanding, and representing VRU objects (pedestrians, cyclists, motorcyclists) and their interactions with other objects. The team owns VRU perception signals and supports VRU applications across all product areas. Our work improves key product metrics like safety, rider comfort, and cost. We also define and execute the roadmap for VRU signals of onboard systems.

In this hybrid role, you will report to the Technical Lead Manager of Semantics Understanding.

You will:

• Develop and deploy deep learning models, including vision language models (VLMs) and Large Language Models (LLMs) for human understanding applications across all product areas
• Define and execute the roadmap of human understanding applications
• Develop data mining, labeling, training and eval pipelines to support the onboard development
• Collaborate and work in partnership with research teams across Alphabet

You have:

• BS degree in Computer Science or a similar discipline, or an equivalent amount of deep learning experience
• 4+ years of industry or research experience with deep Learning, VLM/LLMs, and/or computer vision
• 4+ years of experience with Python
• Hands-on experience with deep learning frameworks such as Jax, PyTorch, Tensorflow

We prefer:

• Experience in autonomous driving and/or robotics industry
• Experience leading projects with >=2 engineers
• Experience applying machine learning in industry problems
• Experience with C++
• Github repositories or Tech Blogs of LLMs/ VLMs
• PhD degree in Computer Science or a similar discipline, or an equivalent amount of deep learning experience