Selected work

Representative Papers

Constrained optimal transport with an application to large markets with indivisible goods

Apr 02, 2026

This study addresses the optimal transport problem between a continuum of agents and a finite set of discrete options under linear constraints, aiming to rectify a flaw in the existence proof of equilibria in large markets with indivisible goods. By developing a Monge–Kantorovich duality theory tailored to this setting and leveraging tools from functional analysis, convex analysis, and potential function optimization, the authors correct a prior erroneous claim regarding compactness. The main contributions are twofold: first, they rigorously reestablish the existence of market equilibria; second, they characterize equilibrium prices as minimizers of a dual potential function, thereby providing a computationally tractable method for equilibrium computation.

0 citationsRead paper

FlowSlider: Training-Free Continuous Image Editing via Fidelity-Steering Decomposition

Apr 02, 2026

Existing image editing methods often rely on additional training and suffer performance degradation under distribution shifts. This work proposes a training-free, continuous editing approach built upon the Rectified Flow framework, which decomposes the editing update into a fidelity term and a guidance term. By enforcing orthogonality between these two components, the method effectively decouples semantic manipulation from image fidelity preservation. The editing strength can be smoothly controlled merely by scaling the guidance term, eliminating the need for any post-hoc training. Extensive experiments demonstrate that the proposed method achieves high-quality, stable, and continuously adjustable editing across diverse tasks, significantly outperforming existing training-dependent approaches.

0 citationsRead paper

DH-EAC: Design of a Dynamic, Hierarchical Entanglement Access Control Protocol

Oct 03, 2025

Existing Dicke-state-based pure quantum multiple-access control (MAC) protocols are confined to static, single-quantum local area networks (QLANs), lack scalability to wide-area, dynamic multi-QLAN topologies, and rely on post-selection coordination. Method: We propose DH-EAC—a pure quantum, dynamic, hierarchical entanglement access control protocol—employing an inner–outer dual-layer quantum lottery mechanism to achieve fair and anonymous entanglement resource allocation without classical communication. Contribution/Results: DH-EAC eliminates round-trip signaling and post-selection by directly determining the winning node set and entitlements via pure quantum measurement. It mitigates allocation skew under heterogeneous network scales, ensuring scalable fairness. Evaluated via Dicke-state MAC modeling, i.i.d. loss characterization, Jain’s fairness index, and end-to-end latency analysis, DH-EAC outperforms both single-layer quantum schemes and classical GO allocators in success probability, throughput, and fairness—delivering a low-latency, highly anonymous, and practically deployable access control solution for multi-QLANs.

0 citationsRead paper

Transtiff: A Stylus-shaped Interface for Rendering Perceived Stiffness of Virtual Objects via Stylus Stiffness Control

Feb 14, 2025

To address stiffness perception distortion of virtual objects during touch interaction in virtual environments, this paper proposes an inverse haptic illusion method based on dynamic modulation of the stylus’s intrinsic stiffness. Unlike conventional approaches relying on virtual rendering or external force feedback, our method is the first to identify and exploit the stylus’s mechanical stiffness as a critical modulator of user-perceived stiffness. We design and implement the first real-time variable-stiffness stylus integrated with McKibben artificial muscles, enabling controllable inverse haptic mapping—e.g., rendering “soft” tactile sensations for physically “hard” virtual objects. By tightly coupling closed-loop force feedback, psychophysical calibration, and synchronized VR rendering, we successfully reproduce perceptual stiffness corresponding to diverse real-world materials (e.g., sponge, plastic, tennis ball). User studies demonstrate a 62% improvement in perceptual fidelity over fixed-stiffness baselines, confirming significant enhancement in cross-modal stiffness congruence.

0 citationsRead paper

Recent publications

Latest Papers

Constrained optimal transport with an application to large markets with indivisible goods

Apr 02, 2026

This study addresses the optimal transport problem between a continuum of agents and a finite set of discrete options under linear constraints, aiming to rectify a flaw in the existence proof of equilibria in large markets with indivisible goods. By developing a Monge–Kantorovich duality theory tailored to this setting and leveraging tools from functional analysis, convex analysis, and potential function optimization, the authors correct a prior erroneous claim regarding compactness. The main contributions are twofold: first, they rigorously reestablish the existence of market equilibria; second, they characterize equilibrium prices as minimizers of a dual potential function, thereby providing a computationally tractable method for equilibrium computation.

0 citationsRead paper

FlowSlider: Training-Free Continuous Image Editing via Fidelity-Steering Decomposition

Apr 02, 2026

Existing image editing methods often rely on additional training and suffer performance degradation under distribution shifts. This work proposes a training-free, continuous editing approach built upon the Rectified Flow framework, which decomposes the editing update into a fidelity term and a guidance term. By enforcing orthogonality between these two components, the method effectively decouples semantic manipulation from image fidelity preservation. The editing strength can be smoothly controlled merely by scaling the guidance term, eliminating the need for any post-hoc training. Extensive experiments demonstrate that the proposed method achieves high-quality, stable, and continuously adjustable editing across diverse tasks, significantly outperforming existing training-dependent approaches.

0 citationsRead paper

DH-EAC: Design of a Dynamic, Hierarchical Entanglement Access Control Protocol

Oct 03, 2025

Existing Dicke-state-based pure quantum multiple-access control (MAC) protocols are confined to static, single-quantum local area networks (QLANs), lack scalability to wide-area, dynamic multi-QLAN topologies, and rely on post-selection coordination. Method: We propose DH-EAC—a pure quantum, dynamic, hierarchical entanglement access control protocol—employing an inner–outer dual-layer quantum lottery mechanism to achieve fair and anonymous entanglement resource allocation without classical communication. Contribution/Results: DH-EAC eliminates round-trip signaling and post-selection by directly determining the winning node set and entitlements via pure quantum measurement. It mitigates allocation skew under heterogeneous network scales, ensuring scalable fairness. Evaluated via Dicke-state MAC modeling, i.i.d. loss characterization, Jain’s fairness index, and end-to-end latency analysis, DH-EAC outperforms both single-layer quantum schemes and classical GO allocators in success probability, throughput, and fairness—delivering a low-latency, highly anonymous, and practically deployable access control solution for multi-QLANs.

0 citationsRead paper

Transtiff: A Stylus-shaped Interface for Rendering Perceived Stiffness of Virtual Objects via Stylus Stiffness Control

Feb 14, 2025

To address stiffness perception distortion of virtual objects during touch interaction in virtual environments, this paper proposes an inverse haptic illusion method based on dynamic modulation of the stylus’s intrinsic stiffness. Unlike conventional approaches relying on virtual rendering or external force feedback, our method is the first to identify and exploit the stylus’s mechanical stiffness as a critical modulator of user-perceived stiffness. We design and implement the first real-time variable-stiffness stylus integrated with McKibben artificial muscles, enabling controllable inverse haptic mapping—e.g., rendering “soft” tactile sensations for physically “hard” virtual objects. By tightly coupling closed-loop force feedback, psychophysical calibration, and synchronized VR rendering, we successfully reproduce perceptual stiffness corresponding to diverse real-world materials (e.g., sponge, plastic, tennis ball). User studies demonstrate a 62% improvement in perceptual fidelity over fixed-stiffness baselines, confirming significant enhancement in cross-modal stiffness congruence.

0 citationsRead paper