Research Intern: Online Mapping and Semantic Scene Understanding for Autonomous Parking

Honda Research Institute USA; Mountain View (CA); On-site; USD 22,000 - 36,000; Part time

Job summary

Honda Research Institute USA in Mountain View, CA, seeks a PhD research intern to advance online mapping and semantic scene understanding for autonomous parking. The project combines onboard camera data with vehicle pose to build a persistent semantic map of parking spaces, aisles, and conditions over time.

The role involves fusing observations, handling uncertainty, benchmarking accuracy, and evaluating map utility for navigation and parking-space selection, with results suitable for

Qualifications

• MS in robotics, CS, or related field.
• Strong background in computer vision, robotic perception, and ML.
• Proficiency in Python and PyTorch.
• Experience implementing and evaluating vision-based methods; ability to conduct independent research.

Responsibilities

• Develop online semantic maps of parking spaces, aisles, entrances, connectivity, and obstructions.
• Fuse observations over time and represent uncertainty, occlusion, and changing conditions.
• Benchmark accuracy, temporal consistency, and runtime using real test-car data.
• Evaluate map utility for parking navigation; document findings, and contribute to publications.

Skills

Python

Deep learning (PyTorch)

Computer vision

Robotics perception

Independent research

Education

MS in Robotics/Computer Science or related field

Tools

ROS/ROS 2

C++

Job description

Research Intern: Online Mapping and Semantic Scene Understanding for Autonomous Parking - Honda Research Institute USA

Job Number: P26INT-01

Honda Research Institute USA (HRI-US) is seeking a Ph.D. research intern to develop online mapping and semantic scene understanding for autonomous parking. The intern will combine onboard camera observations and vehicle pose information to maintain a semantic representation of parking‑lot geometry and conditions over time. The work will support exploration, navigation, and parking‑space selection, with benchmarking using data collected from a real test car.

Mountain View, CA

Key Responsibilities

• Develop online semantic maps of parking spaces, aisles, entrances, connectivity, and obstructions.
• Fuse observations over time and represent uncertainty, occlusion, and changing conditions.
• Benchmark accuracy, temporal consistency, and runtime using real test‑car data.
• Evaluate map utility for parking navigation; document findings, and contribute to publications.

Minimum Qualifications

• M.S. in Robotics, Computer Science, or a related field.
• Strong background in computer vision, robotic perception, and machine learning.
• Proficiency in Python and deep learning frameworks such as PyTorch.
• Experience implementing and evaluating vision‑based methods; ability to conduct independent research.

Bonus Qualifications

• Ph.D. candidate in Robotics, Computer Science, Electrical Engineering, or a related field.
• Experience with semantic mapping, visual SLAM, BEV representations, or temporal sensor fusion.
• Familiarity with autonomous driving datasets, camera calibration, and vehicle pose estimation.
• Experience with C++, ROS/ROS 2, or efficient inference; familiarity with vision‑language models.
• Prior publications in robotics, computer vision, or machine learning.

Years of Work Experience Required

0

Desired Start Date

May 2027

Internship Duration

3 Months

Position Keywords

Semantic mapping, scene understanding, computer vision, autonomous parking, temporal fusion

For California applicants: Please review the Honda California Job Applicant Privacy Notice before submitting your application