Problem
Research questions and friction points this paper is trying to address.
Addresses coarse-fine semantic fusion in segmentation
Proposes progressive learning without extra parameters
Enhances medical image segmentation performance
PL-Net: progressive learning network for medical image segmentation
🤖 AI Summary
To address insufficient fusion of coarse- and fine-grained semantic information in medical image segmentation, this paper proposes PL-Net, a progressive learning network. Methodologically, PL-Net introduces a novel dual-path progressive learning paradigm: (i) Internal Progressive Learning (IPL), which extracts multi-scale receptive field features stepwise within a single forward pass; and (ii) External Progressive Learning (EPL), which sequentially fuses coarse- and fine-grained semantics across training stages—without introducing any additional learnable parameters. Built upon the U-Net architecture, PL-Net integrates multi-scale feature mixing and parameter-free progressive feature decoupling. Evaluated on five mainstream medical segmentation benchmarks, it achieves state-of-the-art or competitive performance while maintaining identical parameter count to standard U-Net. This demonstrates the efficacy of PL-Net’s efficient, lightweight semantic progressive modeling framework.
📝 Abstract
In recent years, deep convolutional neural network-based segmentation methods have achieved state-of-the-art performance for many medical analysis tasks. However, most of these approaches rely on optimizing the U-Net structure or adding new functional modules, which overlooks the complementation and fusion of coarse-grained and fine-grained semantic information. To address these issues, we propose a 2D medical image segmentation framework called Progressive Learning Network (PL-Net), which comprises Internal Progressive Learning (IPL) and External Progressive Learning (EPL). PL-Net offers the following advantages: 1) IPL divides feature extraction into two steps, allowing for the mixing of different size receptive fields and capturing semantic information from coarse to fine granularity without introducing additional parameters; 2) EPL divides the training process into two stages to optimize parameters and facilitate the fusion of coarse-grained information in the first stage and fine-grained information in the second stage. We conducted comprehensive evaluations of our proposed method on five medical image segmentation datasets, and the experimental results demonstrate that PL-Net achieves competitive segmentation performance. It is worth noting that PL-Net does not introduce any additional learnable parameters compared to other U-Net variants.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Progressive Learning Network for medical segmentation
Internal Progressive Learning mixes receptive fields
External Progressive Learning optimizes training stages
🔎 Similar Papers
No similar papers found.
