Deep-Learning-Aided Path Planning and Map Construction for Expediting Indoor Mapping
This addresses the problem of expediting indoor mapping for robotics or autonomous systems, though it appears incremental as it builds on frontier-based planners with a deep learning enhancement.
The paper tackles autonomous indoor mapping by using a pre-trained generative deep neural network as a map predictor in path planning and map construction to minimize mapping time, achieving reductions of over 50% in some cases.
The problem of autonomous indoor mapping is addressed. The goal is to minimize the time to achieve a predefined percentage of exposure with some desired level of certainty. The use of a pre-trained generative deep neural network, acting as a map predictor, in both the path planning and the map construction is proposed in order to expedite the mapping process. This method is examined in combination with several frontier-based path planners for two distinct floorplan datasets. Simulations are run for several configurations of the integrated map predictor, the results of which reveal that by utilizing the prediction a significant reduction in mapping time is possible. When the prediction is integrated in both path planning and map construction processes it is shown that the mapping time may in some cases be cut by over 50%.