Search papers, labs, and topics across Lattice.
This paper introduces the Collaborative Feature Refinement Network (CFR-Net), designed to enhance medical image anomaly detection by addressing the limitations of networks pretrained on natural images. By employing a Multi-Path Feature Refinement Module (MPFRM) that utilizes shared parameters for both teacher and student features, CFR-Net effectively mitigates domain discrepancies and captures fine-grained, multi-scale, and orientation-sensitive features. Experimental results demonstrate that CFR-Net not only achieves competitive anomaly classification but also excels in anomaly localization when trained solely on normal data, highlighting its robustness in medical imaging contexts.
CFR-Net outperforms traditional models in medical image anomaly detection by refining features collaboratively, even when trained only on normal data.
Medical image anomaly detection remains challenging because networks pretrained on natural images often exhibit limited adaptability to medical images, where abnormal patterns appear as fine-grained local shifts, multi-scale contextual mismatches, and orientation-sensitive structural deviations. To address this, we propose the Collaborative Feature Refinement Network (CFR-Net), which combines shared teacher-student feature refinement before decoding with cross-space consistency after decoding. CFR-Net refines frozen teacher features and trainable student features using a Multi-Path Feature Refinement Module (MPFRM) with shared parameters, imposing common multi-path refinement rules on generic visual references and representations adapted to the medical domain, thereby mitigating domain discrepancy while modeling local, multi-scale, and orientation-sensitive feature characteristics. A variance-sensitive objective and dynamic ``homework set'' reorganization further support layer-adaptive consistency learning. Experiments on medical benchmarks show that CFR-Net achieves competitive anomaly classification and strong anomaly localization performance when trained on normal data.
