Toyota Research Institute
Robotics Intern - Large Behavior Models, Generalizable 3D Representations (G3R)
Toyota Research Institute, Los Altos, California, United States, 94024
Robotics Intern - Large Behavior Models, Generalizable 3D Representations (G3R)
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 in Automated Driving, Energy & Materials, Human-Centered AI, Human Interactive Driving, Large Behavior Models, and Robotics. This is a summer 2026 paid 12-week internship opportunity. Please note that this internship will be a hybrid in-office role. As a Research Intern, you will collaborate with a multidisciplinary team proposing, conducting, and transferring pioneering research at the intersection of Computer Vision, Graphics, and Robotics. You will work with large-scale 2D, 3D, and multimodal datasets, design and train new representation learning methods, and validate ideas in both simulation and physical robotic systems. Your work will aim toward high-impact research publications and contributions to TRI's mission of enabling robots that can seamlessly perceive and adapt to the real world. The Generalizable 3D Representations (G3R) team in the Robotics division is looking for research interns for Summer 2026. We are advancing core 3D perception and representation learning methods, enabling robots to understand, reconstruct, and interact with the world across diverse environments and sensor modalities. Our work spans 3D Foundation Models, Generative Modeling for 3D/4D, Scene and Object Reconstruction, Multi-Modal Policy Learning, and Embodied World Modeling. Our mission is to develop generalizable, data-driven 3D representations that bridge computer vision, graphics, and roboticscapable of adapting to novel environments, capturing the structure of the world, and supporting robust downstream applications in perception, planning, and interaction. Responsibilities
Conduct daring research in 3D perception, reconstruction, and representation learning, solving open problems of high theoretical and practical value. Push the boundaries of knowledge in 3D Foundation Models for robotic applications. Collaborate with research scientists and engineers across the G3R team, the Robotics division, TRI, Toyota, and our university partners. Stay up to date with state-of-the-art machine learning, 3D vision, and robotics research. Present your work at internal meetings, top-tier conferences, and through contributions to the broader research community. Qualifications
Currently pursuing a Ph.D. in Machine Learning, Computer Vision, Graphics, Robotics, or related fields. Publication record or strong interest in publishing at high-impact venues (e.g., CVPR, ICCV, ECCV, NeurIPS, ICLR, ICML, CoRL, RSS, ICRA, IROS, SIGGRAPH, RA-L). Passion for large-scale challenges in 3D/4D perception, multimodal foundation models, and physical-world-grounded AI. Proficiency with one or more coding languages and systems, preferably Python, Unix, and a Deep Learning framework (e.g., PyTorch). Ability to collaborate with researchers and engineers to invent and implement novel research ideas. A reliable teammate who loves to think big, dive deep, and deliver with integrity.
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 in Automated Driving, Energy & Materials, Human-Centered AI, Human Interactive Driving, Large Behavior Models, and Robotics. This is a summer 2026 paid 12-week internship opportunity. Please note that this internship will be a hybrid in-office role. As a Research Intern, you will collaborate with a multidisciplinary team proposing, conducting, and transferring pioneering research at the intersection of Computer Vision, Graphics, and Robotics. You will work with large-scale 2D, 3D, and multimodal datasets, design and train new representation learning methods, and validate ideas in both simulation and physical robotic systems. Your work will aim toward high-impact research publications and contributions to TRI's mission of enabling robots that can seamlessly perceive and adapt to the real world. The Generalizable 3D Representations (G3R) team in the Robotics division is looking for research interns for Summer 2026. We are advancing core 3D perception and representation learning methods, enabling robots to understand, reconstruct, and interact with the world across diverse environments and sensor modalities. Our work spans 3D Foundation Models, Generative Modeling for 3D/4D, Scene and Object Reconstruction, Multi-Modal Policy Learning, and Embodied World Modeling. Our mission is to develop generalizable, data-driven 3D representations that bridge computer vision, graphics, and roboticscapable of adapting to novel environments, capturing the structure of the world, and supporting robust downstream applications in perception, planning, and interaction. Responsibilities
Conduct daring research in 3D perception, reconstruction, and representation learning, solving open problems of high theoretical and practical value. Push the boundaries of knowledge in 3D Foundation Models for robotic applications. Collaborate with research scientists and engineers across the G3R team, the Robotics division, TRI, Toyota, and our university partners. Stay up to date with state-of-the-art machine learning, 3D vision, and robotics research. Present your work at internal meetings, top-tier conferences, and through contributions to the broader research community. Qualifications
Currently pursuing a Ph.D. in Machine Learning, Computer Vision, Graphics, Robotics, or related fields. Publication record or strong interest in publishing at high-impact venues (e.g., CVPR, ICCV, ECCV, NeurIPS, ICLR, ICML, CoRL, RSS, ICRA, IROS, SIGGRAPH, RA-L). Passion for large-scale challenges in 3D/4D perception, multimodal foundation models, and physical-world-grounded AI. Proficiency with one or more coding languages and systems, preferably Python, Unix, and a Deep Learning framework (e.g., PyTorch). Ability to collaborate with researchers and engineers to invent and implement novel research ideas. A reliable teammate who loves to think big, dive deep, and deliver with integrity.