Learning Deep Generative Spatial Models for Mobile Robots
This addresses the challenge of fragmented modeling for mobile robots, offering a unified approach for tasks like semantic classification and novelty detection, though it is incremental in combining existing techniques like SPNs and deep learning.
The paper tackles the problem of mobile robots needing separate models for different spatial tasks by proposing a single, universal deep generative model that learns a joint distribution over geometry and semantics, achieving superior performance to task-specific state-of-the-art models like GANs and SVMs in experiments on laser-range data.
We propose a new probabilistic framework that allows mobile robots to autonomously learn deep, generative models of their environments that span multiple levels of abstraction. Unlike traditional approaches that combine engineered models for low-level features, geometry, and semantics, our approach leverages recent advances in Sum-Product Networks (SPNs) and deep learning to learn a single, universal model of the robot's spatial environment. Our model is fully probabilistic and generative, and represents a joint distribution over spatial information ranging from low-level geometry to semantic interpretations. Once learned, it is capable of solving a wide range of tasks: from semantic classification of places, uncertainty estimation, and novelty detection, to generation of place appearances based on semantic information and prediction of missing data in partial observations. Experiments on laser-range data from a mobile robot show that the proposed universal model obtains performance superior to state-of-the-art models fine-tuned to one specific task, such as Generative Adversarial Networks (GANs) or SVMs.