Machine Learning Intern, Perception (End-To-End)

Mission Summary: We are looking for a Machine Learning Research Intern to work on end-to-end autonomous driving, with a focus on learning-based models that connect perception, reasoning, prediction, and action. The research direction may include Vision-Action models, Vision-Language-Action models, world models, world-action models, and other foundation-model-inspired approaches for autonomous driving. This role involves literature review, model prototyping, experiment design, evaluation, and analysis. The internship will take place in our Singapore office and we expect a full-time internship period of at least 5 months. We offer flexible working hours and allow for remote work. The candidate will however have to live in Singapore for the duration of the internship. Responsibilities: Conduct research on end-to-end autonomous driving models, including Vision-Action models, Vision-Language-Action models, world models, world-action models, and related approaches Explore learning-based approaches that connect perception, scene understanding, future prediction, decision-making, and driving action generation Prototype, train, and evaluate models using multi-modal autonomous driving data, including images, videos, LiDAR, radar, maps, ego-motion, trajectories, and driving logs Investigate different learning paradigms, such as imitation learning, reinforcement learning, generative modeling, or hybrid learning-based planning Analyze model performance, robustness, generalization, and failure cases in complex driving scenarios Document research ideas, model designs, experiment results, and key findings in clear and reproducible formats Contribute to technical reports, invention disclosures, patent applications, and research paper writing and submission Required Skills: Currently pursuing a Master’s or PhD degree in Computer Science, Machine Learning, Robotics, Electrical Engineering, or a related field Research experience in at least one of the following areas: computer vision, sequence modeling, imitation learning, reinforcement learning, robotics, planning, or autonomous driving Proficient in Python and/or C++ Proficient with deep learning frameworks such as PyTorch or TensorFlow Experience conducting research-oriented experiments, including model training, evaluation, ablation studies, and result analysis Experience working with structured or unstructured data, such as images, videos, point clouds, trajectories, maps, or sensor logs Strong problem-solving skills and ability to communicate research ideas and experimental findings clearly Preferred Skills: Research or project experience in autonomous driving, robotics, embodied AI, or learning-based planning Experience with end-to-end autonomous driving, Vision-Action models, Vision-Language-Action models, world models, or world-action models Familiarity with imitation learning, reinforcement learning, or generative modeling Familiarity with autonomous driving datasets, benchmarks, or simulators such as nuScenes, Waymo Open Dataset, Argoverse, NAVSIM, CARLA, or related frameworks Experience with large-scale model training, distributed training, or efficient fine-tuning Publications, open-source contributions, or strong research projects in relevant areas are a plus