Spatial Sampling Network for Fast Scene Understanding
This work addresses the need for fast and accurate scene understanding in applications like autonomous driving or robotics, though it appears incremental as it builds upon existing segmentation networks.
The authors tackled efficient scene understanding by proposing a network architecture with novel modules for semantic and instance segmentation, achieving 8.6% higher accuracy than the fastest competitor in semantic segmentation and being nearly five times faster in instance segmentation.
We propose a network architecture to perform efficient scene understanding. This work presents three main novelties: the first is an Improved Guided Upsampling Module that can replace in toto the decoder part in common semantic segmentation networks. Our second contribution is the introduction of a new module based on spatial sampling to perform Instance Segmentation. It provides a very fast instance segmentation, needing only thresholding as post-processing step at inference time. Finally, we propose a novel efficient network design that includes the new modules and test it against different datasets for outdoor scene understanding. To our knowledge, our network is one of the themost efficient architectures for scene understanding published to date, furthermore being 8.6% more accurate than the fastest competitor on semantic segmentation and almost five times faster than the most efficient network for instance segmentation.