Compensating for Context by Learning Local Models of Perception Performance

Perception system performance can vary dramatically with contextual factors such as environmental geometry, appearance, and other phenomena. In this work we present a theoretical framework for understanding the role of context in perception and discuss three approaches for predicting probabilistic performance from observations by efficiently learning local performance models. We compare these approaches with experiments on the monocular and stereo visual odometry systems for a ground robot, and show that they can effectively predict system failures in a wide variety of environments.

Authors

• Humphrey Hu
• George Kantor

Keywords

• Context modeling
• Visual odometry
• Data models
• Training data
• Predictive models
• Prediction algorithms
• Perceptual System
• Wide Variety Of Environments
• Ground Robots
• Artificial Neural Network
• Spatial Information
• Probability Density Function
• K-nearest Neighbor
• Classical Approach
• Density Estimation
• Mixture Model
• Conditional Distribution
• Online Learning
• Noise In Data
• Gaussian Mixture Model
• Non-parametric Approach
• Mobile Robot
• Multiple Environments
• Local Learning
• Distribution Of Performance
• Artificial Neural Network Method
• Velocity Of The Robot
• Logarithmic Standard Deviation
• Parametric Approach
• Image Pairs
• Joint Distribution