Hui Sun

ORCID: 0000-0003-0290-142X
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Algorithms and Data Compression
  • Advanced Computational Techniques and Applications
  • Genomics and Phylogenetic Studies
  • Advanced Data Storage Technologies
  • Corporate Finance and Governance
  • Crystallography and Radiation Phenomena
  • Advanced X-ray Imaging Techniques
  • Gene expression and cancer classification
  • Advanced Data Compression Techniques
  • Remote-Sensing Image Classification
  • Currency Recognition and Detection
  • Evaluation Methods in Various Fields
  • Translation Studies and Practices
  • Civil and Geotechnical Engineering Research
  • Consumer Market Behavior and Pricing
  • Image Processing and 3D Reconstruction
  • Regional Economic and Spatial Analysis
  • Corruption and Economic Development
  • Chromosomal and Genetic Variations
  • DNA and Biological Computing
  • Magnetic properties of thin films
  • Caching and Content Delivery
  • Mobile and Web Applications
  • Supply Chain and Inventory Management
  • Genomic variations and chromosomal abnormalities

Nanyang Technological University
2025

Nankai University
2023-2025

Baidu (China)
2023-2025

Nanjing University of Posts and Telecommunications
2013

Hunan University of Technology
2012

Dalian Institute of Chemical Physics
2012

Chinese Academy of Sciences
2012

Dalian University of Technology
2011

Jiangsu Ocean University
2009

East China Jiaotong University
2007

10.1109/icassp49660.2025.10889184 article EN ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2025-03-12

10.1109/icassp49660.2025.10887721 article EN ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2025-03-12

Learning-based compression shows competitive ratios for genomics data. It often includes three types of compressors: static, adaptive and semi-adaptive. However, these existing compressors suffer from inferior or throughput, also faces model cold-start problems. To address issues, we propose DeepGeCo, a novel data lossless framework with (s,k)-mer encoding deep neural networks, involving modes (MINI PLUS adaptive, ULTRA semi-adaptive) flexible requirements throughput. In (1) develop BiGRU...

10.1609/aaai.v39i12.33371 article EN Proceedings of the AAAI Conference on Artificial Intelligence 2025-04-11

The quality scores data (QSD) account for 70% in compressed FastQ files obtained from the short and long reads sequencing technologies. Designing effective compressors QSD that counterbalance compression ratio, time cost, memory consumption is essential scenarios such as large-scale genomics sharing long-term backup. This study presents a novel parallel lossless QSD-dedicated algorithm named PQSDC, which fulfills above requirements well. PQSDC based on two core components:...

10.1093/bioinformatics/btae323 article EN cc-by Bioinformatics 2024-05-01

The advancement of long reads sequencing technologies has led to a significant increase in biological big data. Although several reference-free compressors are available for saving data storage space, choosing the suitable one is challenging due shortage thorough and systematic evaluations their lossless compression effectiveness, both dedicated general-purpose. In this study, we performed benchmark examinations on 30 compressors, including 11 specialized 19 general-purpose ones, using 31...

10.1109/dcc58796.2024.00101 article EN 2024-03-19

Genomic sequencing reads compressors are essential for balancing high-throughput short generation speed, large-scale genomic data sharing, and infrastructure storage expenditure. However, most existing rarely utilize big-memory systems duplicative information between diverse files to achieve a higher compression ratio conserving space.We employ as the optimization objective propose optimizer, named PMFFRC, through novelty memory modeling redundant clustering technologies. By cascading in 982...

10.1186/s12859-023-05566-9 article EN cc-by BMC Bioinformatics 2023-11-30

Owing to advancements in deep learning technology, Vision Transformers (ViTs) have demonstrated impressive performance various computer vision tasks. Nonetheless, ViTs still face some challenges, such as high computational complexity and the absence of desirable inductive biases. To alleviate these issues, {the potential advantages combining eagle with are explored. We summarize a Bi-Fovea Visual Interaction (BFVI) structure inspired by unique physiological visual characteristics eyes. A...

10.48550/arxiv.2310.06629 preprint EN other-oa arXiv (Cornell University) 2023-01-01
Coming Soon ...