Postdoctoral Researcher, CAD Generation Machine Learning

At Toyota Research Institute (TRI), we’re on a mission to improve the quality of human life. We’re developing new tools and capabilities to amplify the human experience. To lead this transformative shift in mobility, we’ve built a world-class team advancing the state of the art in AI, robotics, driving, and material sciences.

This postdoc opportunity falls within theHuman-Centered AI Division (HCAI). We are an integrated team of ML researchers, behavior scientists, and human-computer interaction experts. At the core of our work, we aim to support people to make better decisions by leveraging the best of big data, technology, and insights about why we do what we do.

We are seeking a highly motivated and talented postdoctoral researcher for a one-year position to join our Future Product Innovation team and push the envelope on what is possible with Generative AI technologies in the domain of automotive product design. The ideal candidate will have a strong background in CAD Generation and Machine Learning or related fields, with demonstrated interest in design and generative AI. In this project, the postdoc will develop 3D shape generation techniques constrained by design requirements and knowledge with applications to CAD.

Over the course of the project, in addition to Future Product Innovation group meetings, the postdoc will also participate in all Human-Centered AI division meetings and meetings with stakeholders from relevant Toyota business groups. The postdoc will be exposed to how interdisciplinary industrial research works and how cross-functional groups collaborate. Furthermore, the postdoc will engage in strategy discussions about how their research connects to business impact at Toyota.

Responsibilities

In collaboration with the other researchers in the department and the stakeholders, you will

• (1) scope the project to align to the core research efforts;
• (2) be the primary driver of the technical plan (e.g., model development, analysis plan) with regular feedback from mentors;
• (3) execute the project using TRI resources; and
• (4) present the project’s approach and findings in research meetings.

Qualifications

• PhD in Computer Science, Machine Learning or related fields.
• Experience working with CAD data such as B-rep and step-file.
• Experience working with relevant libraries such as pythonocc-core for primitive fitting.
• Experience with multimodal generative models for boundary representations.
• Track record of executing research projects including publications at top venues including but not limited to CVPR, ICCV, TOG, SIGGRAPH, NeurIPS.
• Desire to work on challenging open-ended research projects.
• Demonstrated ability to work autonomously while soliciting feedback.
• Excellent communication and teamwork skills.

Bonus Qualifications

• Experience with developing research software prototypes.
• Experience adapting pre-trained models to specific tasks or domains through fine-tuning or similar techniques.
• Experience with working with CAD development tools, libraries, and/or API.