FloorGenT: Generative Vector Graphic Model of Floor Plans for Robotics
This work addresses the need for efficient floor plan modeling in robotics, though it is incremental as it applies existing sequence modeling techniques to a new domain.
The authors tackled the problem of modeling floor plans for robotics by representing them as sequences of line segments and using an autoregressive attention-based neural network, achieving capabilities in generation, completion, and distance prediction tasks.
Floor plans are the basis of reasoning in and communicating about indoor environments. In this paper, we show that by modelling floor plans as sequences of line segments seen from a particular point of view, recent advances in autoregressive sequence modelling can be leveraged to model and predict floor plans. The line segments are canonicalized and translated to sequence of tokens and an attention-based neural network is used to fit a one-step distribution over next tokens. We fit the network to sequences derived from a set of large-scale floor plans, and demonstrate the capabilities of the model in four scenarios: novel floor plan generation, completion of partially observed floor plans, generation of floor plans from simulated sensor data, and finally, the applicability of a floor plan model in predicting the shortest distance with partial knowledge of the environment.