Depth-guided Free-space Segmentation for a Mobile Robot
This addresses indoor navigation for mobile robots, but appears incremental as it builds on existing techniques like DPT and SegFormer.
The paper tackles indoor free-space segmentation for mobile robots by associating large depth values with navigable regions, using an unsupervised masking technique and fine-tuning a SegFormer model on a custom dataset. The method demonstrates sufficient performance in cluttered scenarios, though no concrete numerical results are provided.
Accurate indoor free-space segmentation is a challenging task due to the complexity and the dynamic nature that indoor environments exhibit. We propose an indoors free-space segmentation method that associates large depth values with navigable regions. Our method leverages an unsupervised masking technique that, using positive instances, generates segmentation labels based on textural homogeneity and depth uniformity. Moreover, we generate superpixels corresponding to areas of higher depth and align them with features extracted from a Dense Prediction Transformer (DPT). Using the estimated free-space masks and the DPT feature representation, a SegFormer model is fine-tuned on our custom-collected indoor dataset. Our experiments demonstrate sufficient performance in intricate scenarios characterized by cluttered obstacles and challenging identification of free space.