Morphology and actuation as inductive biases in robotic hand manipulation
研究通过分析机器人手的运动学和驱动阶段，解决不同设计哲学下的控制难度问题，使用条件任务雅可比矩阵等方法进行评估。
0 citationsRead paper
Selected work
Representative Papers
A Controlled Evaluation of Model Rankings and Input Reliance in Surface Water Segmentation
该研究通过多种方法如配置比较、输入压力测试等，解决了表面水分割模型排名稳定性及输入依赖性评估问题。
0 citationsRead paper
Corpus Augmentation for Sign Language Translation via LLM-Guided Video Stitching
This work addresses the challenge of sign language translation, which suffers from scarce high-quality parallel video-text data and poor generalization on long-tail vocabulary and unseen structures. The authors propose a corpus augmentation method that requires no additional annotations, external videos, or generative models: hand gesture clips are extracted from existing annotated videos, paired with sentences generated by a large language model (LLM), and randomly concatenated to synthesize new RGB video–text pairs. Notably, abrupt visual transitions between segments act as an implicit regularizer, outperforming smooth transitions. Integrating CTC alignment, LLM-guided sentence generation, and multimodal representation transformation, the approach achieves a 2.92 BLEU-4 improvement over the GFSLT-VLP baseline under the same training framework, surpassing the previous state-of-the-art result by 0.98 BLEU-4.
0 citationsRead paper
KAYRA: A Microservice Architecture for AI-Assisted Karyotyping with Cloud and On-Premise Deployment
This study addresses the limitations of conventional karyotype analysis—namely, low efficiency, insufficient automation, and challenges in balancing privacy preservation with flexible clinical deployment. The authors propose the first end-to-end, containerized microservice-based AI-assisted karyotyping system, integrating EfficientNet-B5 with U-Net for semantic segmentation, Mask R-CNN for instance detection, and a ResNet-18 classifier. Innovatively, the system employs a cascaded region-of-interest (ROI) focusing strategy and a human-in-the-loop review workflow, enabling dual-mode deployment on both cloud and local infrastructure. Evaluated on 459 chromosomes, the system achieves a segmentation accuracy of 98.91%, with classification and orientation accuracies of 89.1% and 89.76%, respectively—significantly outperforming traditional methods and existing AI approaches—and attains Technology Readiness Level (TRL) 6.
0 citationsRead paper
The impact of tactile sensor configurations on grasp learning efficiency -- a comparative evaluation in simulation
This study addresses the lack of systematic investigation into how tactile sensor placement and density influence learning efficiency in robotic grasping. For the first time, the authors systematically evaluate six tactile sensor configurations within a multi-physics simulation environment using a dual-simulation setup to assess performance in reinforcement learning–based grasping tasks. The results demonstrate that specific sensor layouts consistently enhance both learning efficiency and grasp stability across varying simulation conditions. These findings offer a generalizable optimization strategy for tactile perception design in robotic hands and prosthetic devices, providing actionable insights for improving dexterous manipulation through informed sensor arrangement.
0 citationsRead paper
Recent publications
Latest Papers
Morphology and actuation as inductive biases in robotic hand manipulation
研究通过分析机器人手的运动学和驱动阶段，解决不同设计哲学下的控制难度问题，使用条件任务雅可比矩阵等方法进行评估。
0 citationsRead paper
A Controlled Evaluation of Model Rankings and Input Reliance in Surface Water Segmentation
该研究通过多种方法如配置比较、输入压力测试等，解决了表面水分割模型排名稳定性及输入依赖性评估问题。
0 citationsRead paper
Corpus Augmentation for Sign Language Translation via LLM-Guided Video Stitching
This work addresses the challenge of sign language translation, which suffers from scarce high-quality parallel video-text data and poor generalization on long-tail vocabulary and unseen structures. The authors propose a corpus augmentation method that requires no additional annotations, external videos, or generative models: hand gesture clips are extracted from existing annotated videos, paired with sentences generated by a large language model (LLM), and randomly concatenated to synthesize new RGB video–text pairs. Notably, abrupt visual transitions between segments act as an implicit regularizer, outperforming smooth transitions. Integrating CTC alignment, LLM-guided sentence generation, and multimodal representation transformation, the approach achieves a 2.92 BLEU-4 improvement over the GFSLT-VLP baseline under the same training framework, surpassing the previous state-of-the-art result by 0.98 BLEU-4.
0 citationsRead paper
KAYRA: A Microservice Architecture for AI-Assisted Karyotyping with Cloud and On-Premise Deployment
This study addresses the limitations of conventional karyotype analysis—namely, low efficiency, insufficient automation, and challenges in balancing privacy preservation with flexible clinical deployment. The authors propose the first end-to-end, containerized microservice-based AI-assisted karyotyping system, integrating EfficientNet-B5 with U-Net for semantic segmentation, Mask R-CNN for instance detection, and a ResNet-18 classifier. Innovatively, the system employs a cascaded region-of-interest (ROI) focusing strategy and a human-in-the-loop review workflow, enabling dual-mode deployment on both cloud and local infrastructure. Evaluated on 459 chromosomes, the system achieves a segmentation accuracy of 98.91%, with classification and orientation accuracies of 89.1% and 89.76%, respectively—significantly outperforming traditional methods and existing AI approaches—and attains Technology Readiness Level (TRL) 6.
0 citationsRead paper
The impact of tactile sensor configurations on grasp learning efficiency -- a comparative evaluation in simulation
This study addresses the lack of systematic investigation into how tactile sensor placement and density influence learning efficiency in robotic grasping. For the first time, the authors systematically evaluate six tactile sensor configurations within a multi-physics simulation environment using a dual-simulation setup to assess performance in reinforcement learning–based grasping tasks. The results demonstrate that specific sensor layouts consistently enhance both learning efficiency and grasp stability across varying simulation conditions. These findings offer a generalizable optimization strategy for tactile perception design in robotic hands and prosthetic devices, providing actionable insights for improving dexterous manipulation through informed sensor arrangement.
0 citationsRead paper