U2RLE: Uncertainty-Guided 2-Stage Room Layout Estimation
This work addresses a specific bottleneck in room layout estimation for computer vision applications, offering incremental improvements in handling distant boundaries.
The paper tackles the problem of inaccurate distant floor-wall boundary estimation in room layout estimation by proposing U2RLE, a two-stage CNN architecture that uses uncertainty guidance and a distance-aware loss, resulting in improved performance over state-of-the-art methods on ZInD and Structure3D datasets, particularly for distant walls.
While the existing deep learning-based room layout estimation techniques demonstrate good overall accuracy, they are less effective for distant floor-wall boundary. To tackle this problem, we propose a novel uncertainty-guided approach for layout boundary estimation introducing new two-stage CNN architecture termed U2RLE. The initial stage predicts both floor-wall boundary and its uncertainty and is followed by the refinement of boundaries with high positional uncertainty using a different, distance-aware loss. Finally, outputs from the two stages are merged to produce the room layout. Experiments using ZInD and Structure3D datasets show that U2RLE improves over current state-of-the-art, being able to handle both near and far walls better. In particular, U2RLE outperforms current state-of-the-art techniques for the most distant walls.