A5060023404- Natural Language Processing Techniques
- Topic Modeling
- Speech Recognition and Synthesis
- Text Readability and Simplification
- Robotics and Sensor-Based Localization
- Robot Manipulation and Learning
- Soil Geostatistics and Mapping
- Speech and dialogue systems
- Geochemistry and Geologic Mapping
- 3D Surveying and Cultural Heritage
- Robotic Path Planning Algorithms
- Spectroscopy and Chemometric Analyses
- Multimodal Machine Learning Applications
- RNA and protein synthesis mechanisms
- Microbial Metabolic Engineering and Bioproduction
- Remote-Sensing Image Classification
- Axial and Atropisomeric Chirality Synthesis
- Molecular spectroscopy and chirality
- Human Pose and Action Recognition
- Advanced Image and Video Retrieval Techniques
- Proteins in Food Systems
- Biomedical Text Mining and Ontologies
- Remote Sensing in Agriculture
- Polysaccharides Composition and Applications
- Surgical Simulation and Training
Zhejiang University
2020-2025
Huawei Technologies (China)
2020-2024
Chinese University of Hong Kong
2024
Huawei Technologies (United States)
2022-2024
Chengdu University of Technology
2024
State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation
2024
Tianjin Institute of Industrial Biotechnology
2024
Chinese Academy of Sciences
2022-2024
South China University of Technology
2024
Institute of Infection and Immunity
2024
Promoters are among the most-important and most-basic tools for control of metabolic pathways. However, previous research mainly focused on screening characterization some native promoters in Bacillus subtilis. To develop a broadly applicable promoter system this important platform organism, we created de novo synthetic library (SPL) based consensus sequences by analyzing microarray transcriptome data B. subtilis 168. A total 214 potential spanning gradient strengths was isolated...
Deep learning is characterized by its strong ability of data feature extraction. This method can provide unique advantages when applying it to visible and near-infrared spectroscopy for predicting soil organic matter (SOM) content in those cases where the SOM negatively correlated with spectral reflectance soil. study relied on 248 red samples their 400–2450 nm Fengxin County, Jiangxi Province (China) meet three objectives. First, a multilayer perceptron two convolutional neural networks...
Abstract Background Metabolic engineering has expanded from a focus on designs requiring small number of genetic modifications to increasingly complex driven by advances in multiplex genome editing technologies. However, simultaneously modulating multiple genes the chromosome remains challenging Bacillus subtilis . Thus, developing an efficient and convenient method for B. is imperative. Results Here, we developed CRISPR/Cas9n-based system iterative This enabled us introduce various types...
The types, occurrence and composition of authigenic clay minerals in argillaceous limestone sepiolite-bearing strata the first member Middle Permian Maokou Formation (Mao-1 Member) eastern Sichuan Basin were investigated through outcrop section measurement, core observation, thin identification, argon ion polishing, X-ray diffraction, scanning electron microscope, energy spectrum analysis laser ablation-inductively coupled plasma-mass spectrometry. diagenetic evolution sequence was...
Zongyao Li, Zhanglin Wu, Zhiqiang Rao, Xie YuHao, Guo JiaXin, Daimeng Wei, Hengchao Shang, Wang Minghan, Xiaoyu Chen, Zhengzhe Yu, Li ShaoJun, Lei LiZhi, Hao Yang. Proceedings of the 20th International Conference on Spoken Language Translation (IWSLT 2023). 2023.
Abstract Visible–near‐infrared (Vis‐NIR) spectroscopy is increasingly used to predict soil organic carbon (SOC) content. However, the prediction accuracy of this technology dependent on model selection and study scale. This explored roles spectral variable stratified calibration based type in Vis‐NIR for predicting SOC content at a provincial A total 490 samples, collected Jiangxi Province (southeast China), were modeling with partial least squares regression, support vector machine, random...
Retrieval augmented generation (RAG) has shown promise for enhancing natural language understanding (NLU) capabilities of large models (LLMs) by retrieving relevant knowledge as prompts. Extending RAG to spoken (SLU) represents an important research direction. This paper proposes a approach improving SLU. First, the encoder pretrained automatic speech recognition model is utilized retrieval over training set. The corresponding texts and intent labels are then formulated prompts guide SLU...
Although biocatalysis has garnered widespread attention in both industrial and academic realms, the enzymatic synthesis of chiral oxetanes remains an underdeveloped field. Halohydrin dehalogenases (HHDHs) are industrially relevant enzymes that have been engineered to accomplish reversible transformation epoxides. In our work, a biocatalytic platform was constructed for stereoselective kinetic resolution formation 1,3-disubstituted alcohols. HheC from Agrobacterium radiobacter AD1 identify...
Zhanglin Wu, Zongyao Li, Daimeng Wei, Hengchao Shang, Jiaxin Guo, Xiaoyu Chen, Zhiqiang Rao, Zhengzhe Yu, Jinlong Yang, Shaojun Yuhao Xie, Bin Jiawei Zheng, Ming Zhu, Lizhi Lei, Hao Yanfei Jiang. Proceedings of the 20th International Conference on Spoken Language Translation (IWSLT 2023). 2023.
Zhanglin Wu, Daimeng Wei, Zongyao Li, Zhengzhe Yu, Shaojun Xiaoyu Chen, Hengchao Shang, Jiaxin Guo, Yuhao Xie, Lizhi Lei, Hao Yang, Yanfei Jiang. Proceedings of the Eighth Conference on Machine Translation. 2023.
Punctuation restoration can be crucial for the cascade speech translation system. Traditional approaches typically treat it as a sequential tagging problem, predicting which punctuation mark should follow given word. However, this often requires significant computational and storage resources full-stage training or fine-tuning. Our argument is that pre-trained language models (PLMs) directly leverage their learned knowledge generation, making additional unnecessary. In paper, we propose...
Error correction techniques have been used to refine the output sentences from automatic speech recognition (ASR) models and achieve a lower word error rate (WER). Previous works usually adopt end-to-end has strong dependency on Pseudo Paired Data Original Data. But when only pre-training Data, previous negative effect correction. While fine-tuning source side data must be transcribed by well-trained ASR model, which takes lot of time not universal. In this paper, we propose UCorrect, an...
With the deepening of research on SLAM system, possibility cooperative with multi-robots has been proposed. This paper presents a map matching and localization approach considering an aerial-ground system. The proposed aims to help precisely constructed by two independent systems who have large scale variance view points same route eventually enables ground mobile robot localize itself in global given drone. It contains dense mapping Elevation Map software "Metashape", template algorithm,...
This paper describes our work in the WAT 2020 Indic Multilingual Translation Task. We participated all 7 language pairs (En Bn/Hi/Gu/Ml/Mr/Ta/Te) both directions under constrained condition—using only officially provided data. Using transformer as a baseline, Multi->En and En->Multi translation systems achieve best performances. Detailed data filtering domain selection are keys to performance enhancement experiment, with an average improvement of 2.6 BLEU scores for each pair system 4.6...
Objects grasping, that is, bin-picking, is an area almost all operating robots in the industry are ought to perform in. Although a lot of related approaches this field have been proposed, due poor robustness or high resource consumption many existing works, grasping densely piled objects still faces some huge challenges. In work, we develop bin-picking system for safely and adaptively intensively objects. For challenge occlusion scene, leverages Improved DBSCAN first segment point cloud...