PanoRoom: From the Sphere to the 3D Layout
This addresses indoor scene reconstruction for applications like virtual reality or robotics, representing an incremental improvement with novel method elements.
The paper tackles 3D layout estimation from omnidirectional indoor images by proposing a novel fully convolutional network that outputs accurate probability maps, handling occlusions and recovering complex room shapes. It outperforms state-of-the-art methods in both accuracy and speed, though specific numerical gains are not provided in the abstract.
We propose a novel FCN able to work with omnidirectional images that outputs accurate probability maps representing the main structure of indoor scenes, which is able to generalize on different data. Our approach handles occlusions and recovers complex shaped rooms more faithful to the actual shape of the real scenes. We outperform the state of the art not only in accuracy of the 3D models but also in speed.