Institution profile

University of Victoria

Academic institutionnorthamerica · ca
Official website
Research library161linked papers
Opportunities0open roles
Selected work

Representative Papers

Convolutional Fully-Connected Capsule Network (CFC-CapsNet): A Novel and Fast Capsule Network

Jan 24, 2022Journal of Signal Processing Systems

Capsule Networks (CapsNets) suffer from performance degradation, high computational overhead, and excessive parameter counts in complex image classification tasks. To address these limitations, this paper proposes CFC-CapsNet, which introduces a novel Convolutional–Fully Connected (CFC) capsule layer. This hybrid layer replaces conventional dense capsule structures with fewer yet more discriminative capsules, significantly compressing model size while preserving hierarchical spatial modeling capability. By integrating vectorized feature representations with dynamic routing, the method enhances feature representation efficiency. Extensive experiments on CIFAR-10, SVHN, and Fashion-MNIST demonstrate that CFC-CapsNet achieves average accuracy gains of 1.2–2.8% over baseline CapsNets, accelerates training and inference by 2.3×, and reduces parameter count by 37–51%. The proposed architecture thus achieves a favorable trade-off among accuracy, computational efficiency, and model compactness.

6 citationsRead paper

Bankrupting DoS Attackers

May 17, 2022

This paper addresses the severe cost asymmetry between attackers and defenders in Denial-of-Service (DoS) attacks by proposing an economic deterrence mechanism based on dynamic pricing. The server employs a lightweight probabilistic estimator to continuously characterize legitimate traffic features and dynamically prices each incoming request, thereby ensuring that the attacker’s per-unit cost asymptotically exceeds the combined cost incurred by the server and honest users. Theoretical contributions include: (i) the first asymptotically optimal guarantee wherein the attacker’s cost strictly dominates the defender’s; (ii) provably optimal online pricing algorithms for both synchronous and asynchronous adversarial models; and (iii) tight lower bounds linking estimation error to cost growth. Experimental and competitive analysis demonstrates that, under constant-factor estimation error, the server’s total cost grows strictly slower than the attacker’s—significantly enhancing the economic sustainability and deterrent efficacy of DoS defense.

1 citationsRead paper
Recent publications

Latest Papers