Online, self-supervised terrain classification via discriminatively trained submodular Markov random fields

The authors present a novel approach to the task of autonomous terrain classification based on structured prediction. We consider the problem of learning a classifier that will accurately segment an image into “obstacle” and “ground” patches based on supervised input. Previous approaches to this problem have focused mostly on local appearance; typically, a classifier is trained and evaluated on a pixel-bypixel basis, making an implicit assumption of independence in local pixel neighborhoods. We relax this assumption by modeling correlations between pixels in the submodular MRF framework. We show how both the learning and inference tasks can be simply and efficiently implemented-exact inference via an efficient max flow computation; and learning, via an averaged-subgradient method. Unlike most comparable MRFbased approaches, our method is suitable for implementation on a robot in real-time. Experimental results are shown that demonstrate a marked increase in classification accuracy over standard methods in addition to real-time performance.

Authors

• Paul Vernaza
• Ben Taskar
• Daniel D. Lee

Keywords

• Markov random fields
• Navigation
• Robot sensing systems
• Robot vision systems
• Cameras
• Mobile robots
• Humans
• Stereo vision
• Training data
• Graphical models
• Markov Random Field
• Terrain Classification
• Classification Accuracy
• Gradient Method
• Flow Algorithm
• Inference Task
• Increase In Classification Accuracy
• Training Set
• Least Squares Regression
• Parametrized
• Weight Vector
• Graphical Model
• Simple Solution
• Ground Plane
• Efficient Learning
• Mobile Robot
• Posterior Mode
• Node Features
• Bayesian Regression
• Maximum A Posteriori
• Subgradient Method
• Probability Of Assignment
• Minimum Cut
• Stereo Camera
• Efficient Inference
• Binary Label
• Conditional Random Field
• Step In This Direction
• Linear Classifier
• Random Variables