Ajay Sridhar
Papers
8
Total Citations
234
H-Index
5
About
Ajay Sridhar is a robotics researcher whose work sits at the intersection of machine learning, computer vision, and autonomous navigation. His research focuses on developing generalizable, data-driven policies that enable robots to navigate complex real-world environments — including unfamiliar spaces and crowded social settings — without requiring task-specific programming from scratch. Sridhar's most impactful contribution, **NoMaD** (96 citations), introduces a unified diffusion-based policy framework that elegantly handles both goal-directed navigation and open-ended exploration within a single model. His work on **GNM** (71 citations) demonstrated that training across heterogeneous robot platforms could yield powerful, transferable navigation models — a significant step toward generalizable robotics. Complementing this, **ViNT** extends the foundation model paradigm to visual navigation, while **SACSoN** (31 citations) addresses the nuanced challenge of socially compliant robot behavior in human-populated environments. Through frameworks like **ExAug** and **SELFI**, Sridhar has also advanced techniques for data augmentation and autonomous self-improvement, tackling the persistent scarcity of diverse robotic training data. With over 230 cumulative citations across recent publications, his contributions are shaping the future of scalable, adaptable robotic navigation systems.
Research Focus
Key Achievements
Top Papers
- 1NoMaD: Goal Masked Diffusion Policies for Navigation and Exploration96 citations · 2024
- 2GNM: A General Navigation Model to Drive Any Robot71 citations · 2023
- 3SACSoN: Scalable Autonomous Control for Social Navigation31 citations · 2023
- 4
- 5ViNT: A Foundation Model for Visual Navigation15 citations · 2023
- 6
- 7NoMaD: Goal Masked Diffusion Policies for Navigation and Exploration2 citations · 2023
- 8