Results 61 to 70 of about 2,602 (222)
Named Entity Recognition Model Based on k-best Viterbi Decoupling Knowledge Distillation [PDF]
Jisuanji kexue yu tansuoKnowledge distillation is a general approach to improve the performance of the named entity recognition (NER) models. However, the classical knowledge distillation loss functions are coupled, which leads to poor logit distillation.
ZHAO Honglei, TANG Huanling, ZHANG Yu, SUN Xueyuan, LU Mingyu
doaj +1 more source
Exploring the protein universe with distant similarity detection methods
Protein Science, Volume 35, Issue 1, January 2026.Abstract During the last few years, the body of data on proteins is expanding almost exponentially with the development of advanced methods for gene sequencing, protein structure determination, particularly by cryoelectron microscopy, and structure prediction using artificial intelligence‐based approaches.
Harutyun Sahakyan +3 more
wiley +1 more source
Research of fast decoding for longer constraint length convolutional codes
Tongxin xuebao, 2010 To overcome the disadvantage of Viterbi decoding algorithm, in which its complexity exponentially increases with the increasing constraint length of convolutional codes, and the decoding delay was too large to fit the decoding of longer constraint length
HUANG Xiao-ling, YANG Hua-long
doaj +2 more sources
SYNCHRONIZATION RECOVERY WHEN DECODING MESSAGES USING A MODIFIED VITERBI DECODING ALGORITHM
Известия высших учебных заведений. Поволжский регион:Технические науки, 2020 Background. Modern data transmission systems using cryptographic methods of information protection require high reliability and quality of information transmission. The object of research is protected data transmission systems. The subject of research
Yu. Yu. Sinitsyn, A. B. Sizonenko
doaj +1 more source
High performance ACS for Viterbi decoder using pipeline T-Algorithm
Alexandria Engineering Journal, 2015 Viterbi algorithm is the most popular algorithm used to decode the convolution code, but its computational complexity increases exponentially with the increasing constraint length due to the large number of Trellis transitions.
D. Vaithiyanathan +2 more
doaj +1 more source
Entropy, 2020 A new class of convolutional codes, called skew convolutional codes, that extends the class of classical fixed convolutional codes, is proposed.
Vladimir Sidorenko +3 more
doaj +1 more source
Comparing GFDM With OFDM for Use in 6G Communication From a New Perspective
Journal of Electrical and Computer Engineering, Volume 2026, Issue 1, 2026.The most important challenges facing 6G communication systems are the large amount of data and low latency as compared to previous generations. Therefore, it has become necessary to use advanced technologies to meet the requirements of 6G wireless communication systems.
Mohammed Azher Therib +2 more
wiley +1 more source
Estimating Environmental Transmission Risk From Host Movement Data
Ecology and Evolution, Volume 15, Issue 12, December 2025.We present a comparative modeling framework that integrates host movement, pathogen persistence, and behavior to quantify environmentally mediated transmission. Using wild pig GPS data and SEIR simulations, we show that contact definitions significantly alter transmission dynamics, highlighting the importance of mechanistically defining indirect ...
Himel Talukder +6 more
wiley +1 more source
Engineering Reports, Volume 7, Issue 10, October 2025.This study proposes CRIBC, an advanced Named Entity Recognition (NER) model tailored for Chinese medical text processing. By integrating Chinese‐RoBERTa‐WWM‐EXT with BiLSTM‐CRF and IDCNN, CRIBC enables more accurate entity extraction and knowledge representation.
Yiqing Xu +2 more
wiley +1 more source
Unified Neural Lexical Analysis Via Two‐Stage Span Tagging
CAAI Transactions on Intelligence Technology, Volume 10, Issue 4, Page 1254-1267, August 2025.ABSTRACT Lexical analysis is a fundamental task in natural language processing, which involves several subtasks, such as word segmentation (WS), part‐of‐speech (POS) tagging, and named entity recognition (NER). Recent works have shown that taking advantage of relatedness between these subtasks can be beneficial.
Yantuan Xian +5 more
wiley +1 more source