At present,deep learning has been well applied in many fields.However,due to the high complexity of hypothesis space,numerous training samples are usually required to ensure the reliability of minimizing experience ri...At present,deep learning has been well applied in many fields.However,due to the high complexity of hypothesis space,numerous training samples are usually required to ensure the reliability of minimizing experience risk.Therefore,training a classifier with a small number of training examples is a challenging task.From a biological point of view,based on the assumption that rich prior knowledge and analogical association should enable human beings to quickly distinguish novel things from a few or even one example,we proposed a dynamic analogical association algorithm to make the model use only a few labeled samples for classification.To be specific,the algorithm search for knowledge structures similar to existing tasks in prior knowledge based on manifold matching,and combine sampling distributions to generate offsets instead of two sample points,thereby ensuring high confidence and significant contribution to the classification.The comparative results on two common benchmark datasets substantiate the superiority of the proposed method compared to existing data generation approaches for few-shot learning,and the effectiveness of the algorithm has been proved through ablation experiments.展开更多
Recently,deep learning has achieved great success in visual tracking.The goal of this paper is to review the state-of-the-art tracking methods based on deep learning.First,we categorize the existing deep learning base...Recently,deep learning has achieved great success in visual tracking.The goal of this paper is to review the state-of-the-art tracking methods based on deep learning.First,we categorize the existing deep learning based trackers into three classes according to network structure,network function and network training.For each categorize,we analyze papers in different categories.Then,we conduct extensive experiments to compare the representative methods on the popular OTB-100,TC-128 and VOT2015 benchmarks.Based on our observations.We conclude that:(1)The usage of the convolutional neural network(CNN)model could significantly improve the tracking performance.(2)The trackers with deep features perform much better than those with low-level hand-crafted features.(3)Deep features from different convolutional layers have different characteristics and the effective combination of them usually results in a more robust tracker.(4)The deep visual trackers using end-to-end networks usually perform better than the trackers merely using feature extraction networks.(5)For visual tracking,the most suitable network training method is to per-train networks with video information and online fine-tune them with subsequent observations.Finally,we summarize our manuscript and highlight our insights,and point out the further trends for deep visual tracking.展开更多
Object detection is one of the most fundamental,longstanding and significant problems in the field of computer vision,where detection involves object classification and location.Compared with the traditional object de...Object detection is one of the most fundamental,longstanding and significant problems in the field of computer vision,where detection involves object classification and location.Compared with the traditional object detection algorithms,deep learning makes full use of its powerful feature learning capabilities showing better detection performance.Meanwhile,the emergence of large datasets and tremendous improvement in computer computing power have also contributed to the vigorous development of this field.In the paper,many aspects of generic object detection are introduced and summarized such as traditional object detection algorithms,datasets,evaluation metrics,detection frameworks based on deep learning and state-of-the-art detection results for object detectors.Finally,we discuss several promising directions for future research.展开更多
Nowadays,as the number of textual data is exponentially increasing,sentiment analysis has become one of the most significant tasks in natural language processing(NLP)with increasing attention.Traditional Chinese senti...Nowadays,as the number of textual data is exponentially increasing,sentiment analysis has become one of the most significant tasks in natural language processing(NLP)with increasing attention.Traditional Chinese sentiment analysis algorithms cannot make full use of the order information in context and are inefficient in sentiment inference.In this paper,we systematically reviewed the classic and representative works in sentiment analysis and proposed a simple but efficient optimization.First of all,FastText was trained to get the basic classification model,which can generate pre-trained word vectors as a by-product.Secondly,Bidirectional Long Short-Term Memory Network(Bi-LSTM)utilizes the generated word vectors for training and then merges with FastText to make comprehensive sentiment analysis.By combining FastText and Bi-LSTM,we have developed a new fast sentiment analysis,called FAST-BiLSTM,which consistently achieves a balance between performance and speed.In particular,experimental results based on the real datasets demonstrate that our algorithm can effectively judge sentiments of users’comments,and is superior to the traditional algorithm in time efficiency,accuracy,recall and F1 criteria.展开更多
基金This work was supported by The National Natural Science Foundation of China(No.61402537)Sichuan Science and Technology Program(Nos.2019ZDZX0006,2020YFQ0056)+1 种基金the West Light Foundation of Chinese Academy of Sciences(201899)the Talents by Sichuan provincial Party Committee Organization Department,and Science and Technology Service Network Initiative(KFJ-STS-QYZD-2021-21-001).
文摘At present,deep learning has been well applied in many fields.However,due to the high complexity of hypothesis space,numerous training samples are usually required to ensure the reliability of minimizing experience risk.Therefore,training a classifier with a small number of training examples is a challenging task.From a biological point of view,based on the assumption that rich prior knowledge and analogical association should enable human beings to quickly distinguish novel things from a few or even one example,we proposed a dynamic analogical association algorithm to make the model use only a few labeled samples for classification.To be specific,the algorithm search for knowledge structures similar to existing tasks in prior knowledge based on manifold matching,and combine sampling distributions to generate offsets instead of two sample points,thereby ensuring high confidence and significant contribution to the classification.The comparative results on two common benchmark datasets substantiate the superiority of the proposed method compared to existing data generation approaches for few-shot learning,and the effectiveness of the algorithm has been proved through ablation experiments.
基金the Natural Science Foundation of China no.61802058State Key Laboratory of Novel Software Technology(Nanjing University)no.KFKT2017B17Nanjing University of Information Science and Technology Start-up fund No.2018r057.
文摘Recently,deep learning has achieved great success in visual tracking.The goal of this paper is to review the state-of-the-art tracking methods based on deep learning.First,we categorize the existing deep learning based trackers into three classes according to network structure,network function and network training.For each categorize,we analyze papers in different categories.Then,we conduct extensive experiments to compare the representative methods on the popular OTB-100,TC-128 and VOT2015 benchmarks.Based on our observations.We conclude that:(1)The usage of the convolutional neural network(CNN)model could significantly improve the tracking performance.(2)The trackers with deep features perform much better than those with low-level hand-crafted features.(3)Deep features from different convolutional layers have different characteristics and the effective combination of them usually results in a more robust tracker.(4)The deep visual trackers using end-to-end networks usually perform better than the trackers merely using feature extraction networks.(5)For visual tracking,the most suitable network training method is to per-train networks with video information and online fine-tune them with subsequent observations.Finally,we summarize our manuscript and highlight our insights,and point out the further trends for deep visual tracking.
基金This work is supported in part by the National Natural Science Foundation of China(Grant No.61802058)in part by the International Cooperation and Exchange of the National Natural Science Foundation of China(Grant No.61911530397)+2 种基金in part by the Equipment Advance Research Foundation Project of China(Grant No.61403120106)in part by the Startup Foundation for Introducing Talent of Nanjing University of Information Science and Technology(Grant No.2018r057)in part by the Open Project Program of the State Key Lab of CAD&CG(Grant No.A1919),Zhejiang University,and the PAPD fund.
文摘Object detection is one of the most fundamental,longstanding and significant problems in the field of computer vision,where detection involves object classification and location.Compared with the traditional object detection algorithms,deep learning makes full use of its powerful feature learning capabilities showing better detection performance.Meanwhile,the emergence of large datasets and tremendous improvement in computer computing power have also contributed to the vigorous development of this field.In the paper,many aspects of generic object detection are introduced and summarized such as traditional object detection algorithms,datasets,evaluation metrics,detection frameworks based on deep learning and state-of-the-art detection results for object detectors.Finally,we discuss several promising directions for future research.
基金supported by the National Science Foundation of China(No.61771140)the 2017 Natural Science Foundation of Fujian Provincial Science&Technology Department(No.2018J01560)the 2016 Fujian Education and Scientific Research Project for Young and Middle-aged Teachers(JAT170522).
文摘Nowadays,as the number of textual data is exponentially increasing,sentiment analysis has become one of the most significant tasks in natural language processing(NLP)with increasing attention.Traditional Chinese sentiment analysis algorithms cannot make full use of the order information in context and are inefficient in sentiment inference.In this paper,we systematically reviewed the classic and representative works in sentiment analysis and proposed a simple but efficient optimization.First of all,FastText was trained to get the basic classification model,which can generate pre-trained word vectors as a by-product.Secondly,Bidirectional Long Short-Term Memory Network(Bi-LSTM)utilizes the generated word vectors for training and then merges with FastText to make comprehensive sentiment analysis.By combining FastText and Bi-LSTM,we have developed a new fast sentiment analysis,called FAST-BiLSTM,which consistently achieves a balance between performance and speed.In particular,experimental results based on the real datasets demonstrate that our algorithm can effectively judge sentiments of users’comments,and is superior to the traditional algorithm in time efficiency,accuracy,recall and F1 criteria.