Generative 3D/4D Vision Engineer (Embodied AI)

Nanyang Technological University Singapore; Singapore; On-site; SGD 60,000 - 90,000; Full time

Job summary

NTU Singapore’s College of Computing and Data Science invites applications for a Research Engineer to support the NRFI project on a New Intrinsic-Aware Visual Generation Paradigm. You will develop learning-based systems for generative 3D/4D scene modelling, world models, and embodied intelligence.

The role emphasizes building scalable data pipelines with PyTorch, conducting rigorous experiments, and integrating prototypes in simulation and robotics platforms; strong collaboration and

Qualifications

• Master’s degree in CS, robotics, AI, EE, or related field.
• Strong foundations in ML, DL, CV, and 3D/4D vision or RL or robotics.
• Proficiency in Python and PyTorch; experience with C++, ROS or Isaac Gym is advantageous.
• Experience developing and evaluating research systems, including data pipelines and reproducible experiments.
• Experience in embodied AI, world models, robot learning, sim-to-real transfer, or generative 3D/4D modelling.
• Strong analytical and communication skills; ability to work independently.

Responsibilities

• Develop ML methods for generative 3D/4D scene representations, world models, and embodied AI.
• Build scalable data, training, and evaluation pipelines using PyTorch.
• Design experiments for geometric consistency, temporal dynamics, and physical plausibility.
• Integrate research prototypes in simulation and robotic platforms.
• Document methods and results; contribute to reports and publications.

Skills

Machine learning fundamentals

Computer vision

Python

PyTorch

3D/4D vision

Robotics

C++

ROS

Experiment design

Reproducible research

Education

Master’s degree in CS/Robotics/AI/EE or related field

Tools

ROS

Isaac Gym

Simulation tools

Job description

NTU Singapore’s College of Computing and Data Science invites applications for a Research Engineer to support the NRFI project on a New Intrinsic-Aware Visual Generation Paradigm. You will develop learning-based systems for generative 3D/4D scene modelling, world models, and embodied intelligence.

The role emphasizes building scalable data pipelines with PyTorch, conducting rigorous experiments, and integrating prototypes in simulation and robotics platforms; strong collaboration and