Mariano Phielipp
Papers
5
Total Citations
30
H-Index
2
About
Mariano Phielipp is a robotics and artificial intelligence researcher whose work sits at the intersection of imitation learning, natural language processing, and robot manipulation. His research has pioneered the development of language-conditioned policies that enable robots to interpret and execute human instructions through multimodal inputs combining natural language, vision, and motion demonstration. His most influential contribution, "Language-Conditioned Imitation Learning for Robot Manipulation Tasks" (2020), has garnered 21 citations and addressed a critical gap in human-robot communication by creating richer channels between human experts and robotic systems beyond simple motion trajectories. Building on this foundation, Phielipp has advanced the field further by tackling the practical challenges of training efficiency and cross-robot transferability, developing modular, attention-based architectures that reduce the computational burden of training language-conditioned controllers while enabling their deployment across robots with different dynamics. His earlier work on the MoSART simulation environment also demonstrates a commitment to controls education and research infrastructure. Across his career, Phielipp has consistently pushed toward more generalizable, communicative, and resource-efficient robotic learning systems, making him a notable contributor to the growing field of instruction-following robotics.
Research Focus
Key Achievements
Top Papers
- 1Language-Conditioned Imitation Learning for Robot Manipulation Tasks21 citations · 2020
- 2
- 3Learning modular language-conditioned robot policies through attention2 citations · 2023
- 4
- 5