Selected work

Representative Papers

SAID: Semantic Acoustic Imaging Detector for Sound Event Localization and Detection

Sep 25, 2026

Existing sound source localization models typically estimate only directional information while neglecting spatial regions and energy distributions, thereby limiting their capacity for semantic acoustic imaging. To address this limitation, this work proposes the SAID model, which predicts labeled acoustic maps of active sound sources directly from audio, simultaneously characterizing their spatial extent, energy distribution, and categorical information. Methodologically, we introduce Audio2Sph, a novel pre-trained encoder that enhances audio representations through unsupervised sound energy estimation. Furthermore, a simulation-to-real fine-tuning pipeline is established, integrating self-supervised learning with multi-task optimization. The proposed model achieves first place in the DCASE2026 Task 3 Track A evaluation, attaining a Macro mAP of 0.1080 and a Pearson correlation coefficient of 0.3962.

0 citationsRead paper

Capstan-driven Continuum Surgical Robot: Design, Modeling, and Perception

Aug 13, 2026

This study addresses the challenge of shape and contact force perception in compact spool-driven continuum surgical robots, where direct cable tension measurement is infeasible. To overcome this limitation, the authors propose a co-design framework integrating actuation and sensing. A compliant micro-deformation element embedded within the motor mount enables real-time cable tension feedback. This is combined with a multi-body stubby-beam dynamic model, spatial cable path modeling, and proximal multi-axis force/torque sensing to form a unified perception system. Notably, the approach requires no additional space on the spools, thereby circumventing a key bottleneck in conventional continuum robot sensing. Prototype experiments demonstrate that the system reliably and simultaneously estimates end-effector pose, contact force, and contact location in both single- and double-segment configurations.

0 citationsRead paper

Recent publications

Latest Papers

SAID: Semantic Acoustic Imaging Detector for Sound Event Localization and Detection

Sep 25, 2026

Existing sound source localization models typically estimate only directional information while neglecting spatial regions and energy distributions, thereby limiting their capacity for semantic acoustic imaging. To address this limitation, this work proposes the SAID model, which predicts labeled acoustic maps of active sound sources directly from audio, simultaneously characterizing their spatial extent, energy distribution, and categorical information. Methodologically, we introduce Audio2Sph, a novel pre-trained encoder that enhances audio representations through unsupervised sound energy estimation. Furthermore, a simulation-to-real fine-tuning pipeline is established, integrating self-supervised learning with multi-task optimization. The proposed model achieves first place in the DCASE2026 Task 3 Track A evaluation, attaining a Macro mAP of 0.1080 and a Pearson correlation coefficient of 0.3962.

0 citationsRead paper

Capstan-driven Continuum Surgical Robot: Design, Modeling, and Perception

Aug 13, 2026

This study addresses the challenge of shape and contact force perception in compact spool-driven continuum surgical robots, where direct cable tension measurement is infeasible. To overcome this limitation, the authors propose a co-design framework integrating actuation and sensing. A compliant micro-deformation element embedded within the motor mount enables real-time cable tension feedback. This is combined with a multi-body stubby-beam dynamic model, spatial cable path modeling, and proximal multi-axis force/torque sensing to form a unified perception system. Notably, the approach requires no additional space on the spools, thereby circumventing a key bottleneck in conventional continuum robot sensing. Prototype experiments demonstrate that the system reliably and simultaneously estimates end-effector pose, contact force, and contact location in both single- and double-segment configurations.

0 citationsRead paper