Research Intern (3D/4D)

Our recent related work: Objaverse++

Responsibilities

• Design pipelines to curate or generate 3D/4D datasets focused on embodied perception, egocentric reconstruction, and simulated environments.

• Support research on NeRFs, Gaussian Splatting (3DGS), and embodied reasoning datasets.

• Collaborate with robotics and simulation teams to develop evaluation protocols for 3D spatial reasoning.

• Document datasets thoroughly and contribute to open-source embodied AI benchmarks.

Qualifications

• Strong background in computer science, data engineering, artificial intelligence, or related fields, with hands-on experience in large-scale 3D/4D data systems.

• Proficient in 3D graphics, computer vision, or robotics using tools such as Blender, COLMAP, PyTorch3D, or Open3D.

• Solid understanding of multi-view geometry, SLAM, or NeRF pipelines.

• Strong coding and debugging skills for 3D data preprocessing and optimization.

• (Preferred) Familiarity with Ego4D, Habitat, or other embodied AI datasets and simulators.