ℓ₁DecNet+: A new architecture framework by ℓ₁ decomposition and iteration unfolding for sparse feature segmentation

ℓ₁ based sparse regularization plays a central role in compressive sensing and image processing. In this paper, we propose ℓ₁DecNet, as an unfolded network derived from a variational decomposition model incorporating ℓ₁ related sparse regularization and solved by scaled alternating direction method of multipliers (ADMM). ℓ₁DecNet effectively decomposes an input image into a sparse feature and a learned dense feature, and thus helps the subsequent sparse feature related operations. Based on this, we develop ℓ₁DecNet+, a learnable architecture framework consisting of our ℓ₁DecNet and a segmentation module which operates over extracted sparse features instead of original images. This architecture combines well the benefits of mathematical modeling and data-driven approaches. To our best knowledge, this is the first study to incorporate mathematical image prior into feature extraction in segmentation network structures. Moreover, our ℓ₁DecNet+ framework can be easily extended to 3D case. We evaluate the effectiveness of ℓ₁DecNet+ on two commonly encountered sparse segmentation tasks: retinal vessel segmentation in medical image processing and pavement crack detection in industrial abnormality identification. Experimental results on different datasets demonstrate that, our ℓ₁DecNet+ architecture with various lightweight segmentation modules can achieve equal or better performance than their enlarged versions respectively. This leads to especially practical advantages on resource-limited devices.