Jianjun Yi

Measurement-Efficient Few-Shot Vibration Fault Diagnosis via Physics-Informed Self-Supervised Learning and Adaptive Early Stopping

Zongzhe Ni, Xiancheng Ji, Jianjun Yi, Nuozhou Li et autres

Vibration-based fault diagnosis is widely used for rotating machinery health monitoring, but practical diagnosis is often limited by scarce fault labels and uncertain measurement length. Longer vibration records can improve decision reliability but increase sensing and computational cost, whereas overly short records …

cn (code pays fourni par la source)

Super-LIO: A Robust and Efficient LiDAR-Inertial Odometry System With a Compact Mapping Strategy

Liansheng Wang, Dongjiao He, Jianjun Yi

LiDAR-Inertial Odometry (LIO) is a foundational technique for autonomous systems, yet its deployment on resource-constrained platforms remains challenging due to computational and memory limitations. We propose Super-LIO, a robust LIO system that demands both high performance and accuracy, ideal for applications such …

cn, hk (code pays fourni par la source)

VG2GT: Voxel-Gaussian Splatting Visual Geometry Grounded Transformer

Yibin Zhao, Yihan Pan, Jun Nan, W Z Yang et autres

cn (code pays fourni par la source)

Fault Diagnosis of Hydraulic Systems With an Improved Transition Matrix Hierarchical Network Subject to Multimodal Fusion

Nuozhou Li, Jianjun Yi, Feilong Wang, Hongxing Wang

Abstract To accurately address fault diagnosis problems in hydraulic systems for aerospace component testing, computational methodologies mainly rely on typical neural network structures to design specific models tailored for different types of vibration signals or multimodal sensor data. These methods, though effective …

cn (code pays fourni par la source)

Probabilistic Sampling Networks for Hybrid Structure Planning in Semi-Structured Environments

Xiancheng Ji, Jianjun Yi, Lin Su

The advancement of adaptable industrial robots in intelligent manufacturing is hindered by the inefficiency of traditional motion planning methods in high-dimensional spaces. Therefore, a Dempster-Shafer evidence theory-based hybrid motion planner is proposed, in which a probabilistic sampling network (PSNet) and an enhanced …

cn (code pays fourni par la source)

MGM-LIO: multiscale Gaussian model-based LiDAR-inertial odometry with invariant Kalman filter

Liansheng Wang, Xinke Zhang, Yihan Pan, Jianjun Yi

cn (code pays fourni par la source)

HALO: Hybrid Systolic Arrays via Logical Partitioning for Acceleration of Complex-Valued Neural Networks

Jianjun Yi, Eunbi Jeong, SungHee Yum, J.T. Rhee et autres

Complex-Valued Neural Networks (CVNNs) are an emerging class of deep learning models that process data with both real and imaginary components. By efficiently handling complex-valued representations, CVNNs have gained attention as a promising alternative to traditional Real-Valued Neural Networks (RVNNs), especially in …

kr (code pays fourni par la source)

FSFSplatter: Build Surface and Novel Views with Sparse-Views within 2min

Yibin Zhao, Yihan Pan, Jun Nan, Liwei Chen et autres

Gaussian Splatting has become a leading reconstruction technique, known for its high-quality novel view synthesis and detailed reconstruction. However, most existing methods require dense, calibrated views. Reconstructing from free sparse images often leads to poor surface due to limited overlap and overfitting. …

EJRGF: Efficient Joint Registration of Multiple Point Clouds Using Fast Gaussian Filter

Jianjun Yi, Zhiyong Dai, Yibin Zhao, Liansheng Wang

Joint registration plays a critical role when it comes to aligning multiple point clouds. Despite its capacity to obtain unbiased solutions, current joint registration approaches face substantial computational challenges, particularly regarding processing speed and resource consumption, which impede their practical implementation with …

cn (code pays fourni par la source)

tRF Prospect: tRNA-derived Fragment Target Prediction Based on Neural Network Learning

Daixi Ren, Jianjun Yi, Yongzhen Mo, Mei Yang et autres

Objective Transfer RNA-derived fragments (tRFs) are a recently characterized and rapidly expanding class of small non-coding RNAs, typically ranging from 13 to 50 nucleotides in length. They are derived from mature or precursor tRNA molecules through specific cleavage events and have been …

Multi-LiDAR-Inertial SLAM with Temporally-Coherent Online Calibration

Liansheng Wang, Xinke Zhang, Hangbo Ye, Chaojie Wang et autres

Recent advancements in sensor fusion have expanded the applications of multi-LiDAR systems in localization and mapping. However, integrating heterogeneous sensor observations from diverse modalities and perspectives presents significant challenges for simultaneous localization and mapping (SLAM) systems. This paper proposes a novel continuous-time …

cn (code pays fourni par la source)

LLM-Based Heuristic Task Planning for Robotic Arm Proximity Operations with Non-Cooperative Spacecraft

Ying Jin, Wenchao Huang, Yanyan Wu, Jianjun Yi

When performing operational tasks such as capturing or repairing non-cooperative spacecraft, space robotic arms are typically constrained by incomplete prior knowledge and limited global visual observation conditions, which prevents them from autonomously completing tasks requiring structural reasoning about targets. This study focuses …

cn (code pays fourni par la source)