Estimation, Perception, and Intelligent Computing
The George Washington University
EPIC Lab builds reliable perception and estimation systems for robotics, AR/VR, and physical AI. We are interested in intelligent systems that must operate in the real world under uncertainty, limited compute, and imperfect sensing.
-
State Estimation
Robust and consistent inference for dynamic and uncertain environments -
Visual-Inertial Odometry
Reliable motion estimation for robotics and AR/VR systems -
3D Scene Representation
Geometry-aware models for perception, interaction, and spatial understanding -
Resource-Constrained Intelligence
Methods for edge, embedded, and compute-limited platforms -
Robotics and Physical AI
Perception, planning, and decision-making for embodied systems -
AR/VR and Spatial Computing
Real-time scene understanding for immersive environments
- reading-group
Shared weekly reading group papers for EPIC Lab members
Current and upcoming work includes:
- visual-inertial estimation
- edge and quantized VIO
- 3D Gaussian-based scene modeling
- perception in extreme environments
- embodied perception and physical AI systems
This GitHub organization is used for EPIC Lab code, shared resources, and reading group materials. More repositories will be added as projects are released.