HaarNet: Large-scale Linear-Morphological Hybrid Network for RGB-D Semantic Segmentation
This work addresses semantic segmentation for RGB-D data, offering a novel hybrid approach that could improve efficiency in geometry-based learning tasks, though it appears incremental as it builds on existing morphological and linear methods.
The paper tackled RGB-D semantic segmentation by proposing HaarNet, a hybrid network combining linear and morphological operators to process RGB and depth signals effectively with fewer parameters. Experiments showed it is competitive with a state-of-the-art CNN on large-scale datasets.
Signals from different modalities each have their own combination algebra which affects their sampling processing. RGB is mostly linear; depth is a geometric signal following the operations of mathematical morphology. If a network obtaining RGB-D input has both kinds of operators available in its layers, it should be able to give effective output with fewer parameters. In this paper, morphological elements in conjunction with more familiar linear modules are used to construct a mixed linear-morphological network called HaarNet. This is the first large-scale linear-morphological hybrid, evaluated on a set of sizeable real-world datasets. In the network, morphological Haar sampling is applied to both feature channels in several layers, which splits extreme values and high-frequency information such that both can be processed to improve both modalities. Moreover, morphologically parameterised ReLU is used, and morphologically-sound up-sampling is applied to obtain a full-resolution output. Experiments show that HaarNet is competitive with a state-of-the-art CNN, implying that morphological networks are a promising research direction for geometry-based learning tasks.