Multi-Pass, Multi-View Blended Learning for High-Fidelity Volumetric CT Synthesis from Chest X-Rays
本文通过多通道多视角混合学习框架解决从单张2D胸透重建3D CT的问题，提高结构完整性和解剖细节。
0 citationsRead paper
Selected work
Representative Papers
BreastMammo and DenseMammo: Benchmarks for Mammography Domain Generalization
This study addresses the severe domain shift in mammography caused by equipment from different vendors, which significantly hinders the cross-site generalization of AI models. To tackle this issue, the authors introduce two new datasets, BreastMammo and DenseMammo, and propose a foreground-specific histogram matching protocol tailored for mammographic images. Integrated with a Swin Transformer backbone, this approach establishes the first domain generalization benchmark for breast density classification. Evaluated via five-fold cross-validation and external testing on datasets such as TNMammo and LUMINA, the method achieves an internal AUC of 98.32%, substantially outperforming existing techniques like MixStyle and discrete Fourier transform–based methods. The results demonstrate its effectiveness in mitigating domain shifts arising from clinical source variations and enhancing model robustness across domains.
0 citationsRead paper
Uncertainty-Aware Ensemble Deep Randomized Neural Networks for Classification
This work addresses the limitations of existing deep random vector functional-link networks (dRVFL), which treat all samples equally and thus struggle with real-world data contaminated by noise and outliers, leading to degraded discriminative performance due to error propagation through hidden layers. To overcome this, the study introduces an intuitionistic fuzzy mechanism into dRVFL for the first time, adaptively computing membership and non-membership degrees by jointly modeling each sample’s distance to its class center and the heterogeneity of its local neighborhood. This enables effective differentiation among clean, noisy, and anomalous samples, assigning them differentiated weights accordingly. Combined with kernel-space neighborhood analysis and ensemble learning, the proposed method significantly outperforms current state-of-the-art fuzzy and non-fuzzy approaches on UCI and KEEL benchmark datasets under both Gaussian-noise and noise-free settings, demonstrating superior robustness and generalization capability.
0 citationsRead paper
Communication Security via Temporal Dependency
This work proposes a novel security paradigm grounded in temporal dependency, treating time coupling as a fundamental security resource—a perspective unprecedented in prior literature. Traditional communication security relies on shared secrets or channel advantages, which are often infeasible in infrastructure-less, emergency, or highly dynamic wireless networks. The proposed approach leverages state-chained Random Linear Network Coding (RLNC), integrated with synchronized state embedding, adaptive power control, and optimized intentional interference, to ensure that even if an eavesdropper successfully decodes transmitted symbols, it cannot correctly interpret the underlying data. Operating under a strong security model—requiring no pre-shared keys, no channel advantage, and assuming the adversary possesses full knowledge of the protocol—the scheme induces persistent desynchronization in the eavesdropper within sub-second timescales, rendering its decoding irrecoverably erroneous over the long term.
0 citationsRead paper
Latency Decoupling in Low-Feedback Multi-User Networks via Overhearing-Driven NOMA
This work addresses the latency bottleneck in heterogeneous multi-user systems under low-feedback scenarios—such as non-terrestrial networks and massive IoT—where conventional protocols struggle due to their reliance on frequent feedback and the presence of unknown, asymmetric channels. The authors propose ONOMA, a cross-layer transmission scheme that integrates random linear network coding with symbol-aware non-orthogonal multiple access (NOMA). By leveraging user listening and acknowledgment timing, ONOMA implicitly infers channel strength without requiring channel state information at the transmitter, enabling adaptive power allocation. Furthermore, symbol reconstruction ensures interference-free decoding for strong users, effectively decoupling user latencies. Experimental results demonstrate that ONOMA reduces completion time by up to 34% over TDMA, FDMA, multicast, inter-session coding, and classical NOMA in two-user settings, with gains reaching 50% in large-scale asymmetric networks.
0 citationsRead paper
Recent publications
Latest Papers
Multi-Pass, Multi-View Blended Learning for High-Fidelity Volumetric CT Synthesis from Chest X-Rays
本文通过多通道多视角混合学习框架解决从单张2D胸透重建3D CT的问题，提高结构完整性和解剖细节。
0 citationsRead paper
BreastMammo and DenseMammo: Benchmarks for Mammography Domain Generalization
This study addresses the severe domain shift in mammography caused by equipment from different vendors, which significantly hinders the cross-site generalization of AI models. To tackle this issue, the authors introduce two new datasets, BreastMammo and DenseMammo, and propose a foreground-specific histogram matching protocol tailored for mammographic images. Integrated with a Swin Transformer backbone, this approach establishes the first domain generalization benchmark for breast density classification. Evaluated via five-fold cross-validation and external testing on datasets such as TNMammo and LUMINA, the method achieves an internal AUC of 98.32%, substantially outperforming existing techniques like MixStyle and discrete Fourier transform–based methods. The results demonstrate its effectiveness in mitigating domain shifts arising from clinical source variations and enhancing model robustness across domains.
0 citationsRead paper
Uncertainty-Aware Ensemble Deep Randomized Neural Networks for Classification
This work addresses the limitations of existing deep random vector functional-link networks (dRVFL), which treat all samples equally and thus struggle with real-world data contaminated by noise and outliers, leading to degraded discriminative performance due to error propagation through hidden layers. To overcome this, the study introduces an intuitionistic fuzzy mechanism into dRVFL for the first time, adaptively computing membership and non-membership degrees by jointly modeling each sample’s distance to its class center and the heterogeneity of its local neighborhood. This enables effective differentiation among clean, noisy, and anomalous samples, assigning them differentiated weights accordingly. Combined with kernel-space neighborhood analysis and ensemble learning, the proposed method significantly outperforms current state-of-the-art fuzzy and non-fuzzy approaches on UCI and KEEL benchmark datasets under both Gaussian-noise and noise-free settings, demonstrating superior robustness and generalization capability.
0 citationsRead paper
Communication Security via Temporal Dependency
This work proposes a novel security paradigm grounded in temporal dependency, treating time coupling as a fundamental security resource—a perspective unprecedented in prior literature. Traditional communication security relies on shared secrets or channel advantages, which are often infeasible in infrastructure-less, emergency, or highly dynamic wireless networks. The proposed approach leverages state-chained Random Linear Network Coding (RLNC), integrated with synchronized state embedding, adaptive power control, and optimized intentional interference, to ensure that even if an eavesdropper successfully decodes transmitted symbols, it cannot correctly interpret the underlying data. Operating under a strong security model—requiring no pre-shared keys, no channel advantage, and assuming the adversary possesses full knowledge of the protocol—the scheme induces persistent desynchronization in the eavesdropper within sub-second timescales, rendering its decoding irrecoverably erroneous over the long term.
0 citationsRead paper
Latency Decoupling in Low-Feedback Multi-User Networks via Overhearing-Driven NOMA
This work addresses the latency bottleneck in heterogeneous multi-user systems under low-feedback scenarios—such as non-terrestrial networks and massive IoT—where conventional protocols struggle due to their reliance on frequent feedback and the presence of unknown, asymmetric channels. The authors propose ONOMA, a cross-layer transmission scheme that integrates random linear network coding with symbol-aware non-orthogonal multiple access (NOMA). By leveraging user listening and acknowledgment timing, ONOMA implicitly infers channel strength without requiring channel state information at the transmitter, enabling adaptive power allocation. Furthermore, symbol reconstruction ensures interference-free decoding for strong users, effectively decoupling user latencies. Experimental results demonstrate that ONOMA reduces completion time by up to 34% over TDMA, FDMA, multicast, inter-session coding, and classical NOMA in two-user settings, with gains reaching 50% in large-scale asymmetric networks.
0 citationsRead paper