Pioneering the Future of AGI
Combining theoretical breakthroughs with real-world applications to redefine AI through spatial intelligence.
Key Research Areas
Physical Intelligence
Understanding and modeling physical interactions in the real world.
- Physics-informed neural networks
- Object permanence modeling
- Force dynamics prediction
Spatial Reasoning
Developing models that understand and reason about 3D space.
- Geometric deep learning
- Scene understanding
- Spatial relationships
Neuro-Symbolic Reasoning
Merging symbolic and statistical approaches for robust physical reasoning
- Hybrid architectures for physical understanding
- Causal reasoning in 3D environments
- Symbolic grounding for spatial concepts
Compressed Spatial Representations
Developing efficient methods for scaling real-world reasoning
- Neural compression for 3D data
- Efficient spatial transformers
- Memory-augmented architectures
Multi-Agent Systems
Creating frameworks for collaborative AI systems.
- Distributed learning
- Agent communication
- Collective intelligence
Safety & Robustness
Ensuring reliable and safe AI systems.
- Uncertainty quantification
- Adversarial robustness
- Safety constraints
Latest Publications
3D-LFM: Lifting Foundation Model
Mosam Dabhi, László A. Jeni, Simon Lucey · CVPR 2024
A universal 2D-3D lifting model that processes diverse objects without category-specific knowledge.
Open Research
We believe in advancing the field through open collaboration. Access our datasets, benchmarks, and join us in pushing the boundaries of spatial intelligence.
Explore Resources