Dynamic Dual Sampling Module for Fine-Grained Semantic Segmentation
This work addresses fine-grained segmentation for computer vision applications, but it appears incremental as it builds on existing methods to better explore the interrelationship between semantic context and local details.
The paper tackles the problem of fine-grained semantic segmentation by proposing a Dynamic Dual Sampling Module (DDSM) to dynamically model affinities and propagate semantic context to local details, resulting in improved discriminative representation and well-preserved boundaries, with validation on Cityscapes and Camvid datasets.
Representation of semantic context and local details is the essential issue for building modern semantic segmentation models. However, the interrelationship between semantic context and local details is not well explored in previous works. In this paper, we propose a Dynamic Dual Sampling Module (DDSM) to conduct dynamic affinity modeling and propagate semantic context to local details, which yields a more discriminative representation. Specifically, a dynamic sampling strategy is used to sparsely sample representative pixels and channels in the higher layer, forming adaptive compact support for each pixel and channel in the lower layer. The sampled features with high semantics are aggregated according to the affinities and then propagated to detailed lower-layer features, leading to a fine-grained segmentation result with well-preserved boundaries. Experiment results on both Cityscapes and Camvid datasets validate the effectiveness and efficiency of the proposed approach. Code and models will be available at \url{x3https://github.com/Fantasticarl/DDSM}.