Owing to the wide range of applications in various fields,generative models have become increasingly popular.However,they do not handle spatio-temporal features well.Inspired by the recent advances in these models,thi...Owing to the wide range of applications in various fields,generative models have become increasingly popular.However,they do not handle spatio-temporal features well.Inspired by the recent advances in these models,this paper designs a distributed spatio-temporal generative adversarial network(STGAN-D)that,given some initial data and random noise,generates a consecutive sequence of spatio-temporal samples which have a logical relationship.This paper builds a spatio-temporal discriminator to distinguish whether the samples generated by the generator meet the requirements for time and space coherence,and builds a controller for distributed training of the network gradient updated to separate the model training and parameter updating,to improve the network training rate.The model is trained on the skeletal dataset and the traffic dataset.In contrast to traditional generative adversarial networks(GANs),the proposed STGAN-D can generate logically coherent samples with the corresponding spatial and temporal features while avoiding mode collapse.In addition,this paper shows that the proposed model can generate different styles of spatio-temporal samples given different random noise inputs,and the controller can improve the network training rate.This model will extend the potential range of applications of GANs to areas such as traffic information simulation and multiagent adversarial simulation.展开更多
Due to the volume conduction,electroencephalogram（EEG） gives a rather blurred image of brain activities. It is a challenge for generating satisfactory performance with EEG. This paper studies the multiple areas fusi...Due to the volume conduction,electroencephalogram（EEG） gives a rather blurred image of brain activities. It is a challenge for generating satisfactory performance with EEG. This paper studies the multiple areas fusion of EEG classifiers to improve the motor imagery EEG classification performance. Two feature extraction methods are employed to extract the feature from three different areas of EEG. One is power spectral density（PSD）, and the other is common spatial patterns（CSP）. Classifiers are designed based on the well-known linear discrimination analysis（LDA）. The fusion of the individual classifiers is realized by means of the Choquet fuzzy integral. It is demonstrated that the proposed method comes with better performance compared with the individual classifier.展开更多
基金the National Natural Science Foundation of China(61573285).
文摘Owing to the wide range of applications in various fields,generative models have become increasingly popular.However,they do not handle spatio-temporal features well.Inspired by the recent advances in these models,this paper designs a distributed spatio-temporal generative adversarial network(STGAN-D)that,given some initial data and random noise,generates a consecutive sequence of spatio-temporal samples which have a logical relationship.This paper builds a spatio-temporal discriminator to distinguish whether the samples generated by the generator meet the requirements for time and space coherence,and builds a controller for distributed training of the network gradient updated to separate the model training and parameter updating,to improve the network training rate.The model is trained on the skeletal dataset and the traffic dataset.In contrast to traditional generative adversarial networks(GANs),the proposed STGAN-D can generate logically coherent samples with the corresponding spatial and temporal features while avoiding mode collapse.In addition,this paper shows that the proposed model can generate different styles of spatio-temporal samples given different random noise inputs,and the controller can improve the network training rate.This model will extend the potential range of applications of GANs to areas such as traffic information simulation and multiagent adversarial simulation.
文摘Due to the volume conduction,electroencephalogram（EEG） gives a rather blurred image of brain activities. It is a challenge for generating satisfactory performance with EEG. This paper studies the multiple areas fusion of EEG classifiers to improve the motor imagery EEG classification performance. Two feature extraction methods are employed to extract the feature from three different areas of EEG. One is power spectral density（PSD）, and the other is common spatial patterns（CSP）. Classifiers are designed based on the well-known linear discrimination analysis（LDA）. The fusion of the individual classifiers is realized by means of the Choquet fuzzy integral. It is demonstrated that the proposed method comes with better performance compared with the individual classifier.
