Shulin Feng

AUV: Teaching Audio Universal Vector Quantization with Single Nested Codebook

Yuehua Chen, Kai Hu, Long Zhou, Shulin Feng et autres

cn (code pays fourni par la source)

A Framework for Industrial Equipment State Monitoring Under Time-Varying Conditions

Huixin Tian, Shulin Feng, Yuhang Kang, Hongyi Wang et autres

Industrial equipment operational state monitoring can promptly detect anomalies and prevent faults. However, data-driven models struggle to maintain high precision performance under time-varying conditions. Model update techniques are among the viable solutions, yet they are still encumbered by high resource demands, poor …

cn (code pays fourni par la source)

U-Codec: Ultra Low Frame-rate Neural Speech Codec for Fast High-fidelity Speech Generation

Xusheng Yang, Long Zhou, Wenfu Wang, Kai Hu et autres

We propose \textbf{U-Codec}, an \textbf{U}ltra low frame-rate neural speech \textbf{Codec} that achieves high-fidelity reconstruction and fast speech generation at an extremely low frame-rate of 5Hz (5 frames per second). Extreme compression at 5Hz typically leads to severe intelligibility and spectral detail loss, …

Target tracking based on the extended line of sight guidance law and information fusion

Shulin Feng, Guilin Zhang, Rong Gao, Yuanhua Yang

The extended line of sight (LOS) guidance law is designed to track toward a target in a three-dimensional space, and this method presents a family of guidance laws resulting in a rich behaviour for different parameters. First, polar kinematics models for the …

cn (code pays fourni par la source)

Parallel navigation for 3-D autonomous vehicles

Shulin Feng, Shuning Zhang, Mingming Xu, Guanlong Deng

In this paper, parallel navigation is proposed to track the target in three-dimensional space.Firstly, the polar kinematics models for the vehicle and the target are established.Secondly, parallel navigation is derived by using polar kinematics models.Thirdly, cell decomposition method is applied to implement …

Reciprocating compressor health monitoring based on BSInformer with deep convolutional AutoEncoder

Huixin Tian, Bingtian Ju, Shulin Feng

cn (code pays fourni par la source)

Compressed MoE ASR Model Based on Knowledge Distillation and Quantization

Yuping Yuan, Zhao You, Shulin Feng, Dan Su et autres

3M: Multi-loss, Multi-path and Multi-level Neural Networks for speech recognition

Zhao You, Shulin Feng, Dan Su, Dong Yu

Recently, Conformer based CTC/AED model has become a mainstream architecture for ASR. In this paper, based on our prior work, we identify and integrate several approaches to achieve further improvements for ASR tasks, which we denote as multi-loss, multi-path and multi-level, summarized …

cn, us (code pays fourni par la source)

Wearable Bowel Sound Monitoring with Quality Enhancement using U-net

Kang Zhao, Shulin Feng, Hanjun Jiang, Zhihua Wang et autres

Human bowel sounds (BSs) contain rich information on gastrointestinal health status. Recently, the wearable BS monitoring systems has made it possible to record the human BSs in the long term. However, BSs collected by long-term monitoring devices usually contain various types of …

cn (code pays fourni par la source)

Speechmoe2: Mixture-of-Experts Model with Improved Routing

Zhao You, Shulin Feng, Dan Su, Dong Yu

Mixture-of-experts based acoustic models with dynamic routing mechanisms have proved promising results for speech recognition. The design principle of router architecture is important for the large model capacity and high computational efficiency. Our previous work SpeechMoE only uses local grapheme embedding to …

cn, us (code pays fourni par la source)

3M: Multi-loss, Multi-path and Multi-level Neural Networks for speech recognition

Zhao You, Shulin Feng, Dan Su, Dong Yu

Recently, Conformer based CTC/AED model has become a mainstream architecture for ASR. In this paper, based on our prior work, we identify and integrate several approaches to achieve further improvements for ASR tasks, which we denote as multi-loss, multi-path and multi-level, summarized …

Consensus Protocol to Implement Target Tracking

Rong-Hua Wang, Wenjuan Xia, Shulin Feng

Abstract Consensus protocol is suggested for target following of two non-holonomic robots. For moving robot platform, the computer that acts as the controller, and the robots to establish wireless local area network (WLAN), the remote control of the robot is achieved by …

cn (code pays fourni par la source)