In the video captioning methods based on an encoder-decoder,limited visual features are extracted by an encoder,and a natural sentence of the video content is generated using a decoder.However,this kind ofmethod is de...In the video captioning methods based on an encoder-decoder,limited visual features are extracted by an encoder,and a natural sentence of the video content is generated using a decoder.However,this kind ofmethod is dependent on a single video input source and few visual labels,and there is a problem with semantic alignment between video contents and generated natural sentences,which are not suitable for accurately comprehending and describing the video contents.To address this issue,this paper proposes a video captioning method by semantic topic-guided generation.First,a 3D convolutional neural network is utilized to extract the spatiotemporal features of videos during the encoding.Then,the semantic topics of video data are extracted using the visual labels retrieved from similar video data.In the decoding,a decoder is constructed by combining a novel Enhance-TopK sampling algorithm with a Generative Pre-trained Transformer-2 deep neural network,which decreases the influence of“deviation”in the semantic mapping process between videos and texts by jointly decoding a baseline and semantic topics of video contents.During this process,the designed Enhance-TopK sampling algorithm can alleviate a long-tail problem by dynamically adjusting the probability distribution of the predicted words.Finally,the experiments are conducted on two publicly used Microsoft Research Video Description andMicrosoft Research-Video to Text datasets.The experimental results demonstrate that the proposed method outperforms several state-of-art approaches.Specifically,the performance indicators Bilingual Evaluation Understudy,Metric for Evaluation of Translation with Explicit Ordering,Recall Oriented Understudy for Gisting Evaluation-longest common subsequence,and Consensus-based Image Description Evaluation of the proposed method are improved by 1.2%,0.1%,0.3%,and 2.4% on the Microsoft Research Video Description dataset,and 0.1%,1.0%,0.1%,and 2.8% on the Microsoft Research-Video to Text dataset,respectively,compared with the existing video captioning methods.As a result,the proposed method can generate video captioning that is more closely aligned with human natural language expression habits.展开更多
Why do taxonomists matter? The work of taxonomists is often understated if not completely misunderstood. Without taxonomists, organisms cannot be accurately identified, neither can these organisms be given universally...Why do taxonomists matter? The work of taxonomists is often understated if not completely misunderstood. Without taxonomists, organisms cannot be accurately identified, neither can these organisms be given universally accepted names, and reliably positioned in the phylogenetic tree of life. Thanks to the work of taxonomists over the last 269 years since Carl Linnaeus established the binomial system, we can now measure the health and wealth of our biodiversity in a refined, science-based inventory prescribed by stringent nomenclatural rules.展开更多
Taxonomy plays an important role in understanding the origin, evolution, and ecological functionality of biodiversity. There are large number of unknown species yet to be described by taxonomists, which together with ...Taxonomy plays an important role in understanding the origin, evolution, and ecological functionality of biodiversity. There are large number of unknown species yet to be described by taxonomists, which together with their ecosystem services cannot be effectively protected prior to description. Despite this, taxonomy has been increasingly underrated insufficient funds and permanent positions to retain young talents. Further, the impact factordriven evaluation systems in China exacerbate this downward trend, so alternative evaluation metrics are urgently necessary. When the current generation of outstanding taxonomists retires,there will be too few remaining taxonomists left to train the next generation. In light of these challenges, all co-authors worked together on this paper to analyze the current situation of taxonomy and put out a joint call for immediate actions to advance taxonomy in China.展开更多
Modification is one of the most important and effective methods to improve the photoelectrocatalytic(PEC)performance of ZnO.In this paper,the Ru_(x)Zn_(1-x)O/Ti electrodes were prepared by thermal decomposition method...Modification is one of the most important and effective methods to improve the photoelectrocatalytic(PEC)performance of ZnO.In this paper,the Ru_(x)Zn_(1-x)O/Ti electrodes were prepared by thermal decomposition method and the effect of Ru content on those electrodes' electronic structure was analyzed through the first-principles calculation.Various tests were also performed to observe the microstructures and PEC performance.The results showed that as the Ru^(4+) transferred into ZnO lattice and replaced a number of Zn^(2+),the conduction band of ZnO moved downward and the valence band went upward.The number of photogenerated electron-hole pairs increased as the impurity levels appeared in the band gap.In addition,ZnO nanorods exhibited a smaller grain size and a rougher surface under the effect of Ru.Meanwhile,the RuO_(2) nanoparticles on the surface of ZnO nanorods acted as the electron-transfer channel,helping electrons transfer to the counter electrode and delaying the recombination of the electron-hole pairs.Specifically,the Ru_(x)Zn_(1-x)O/Ti electrodes with 9.375 mol% Ru exhibited the best PEC performance with a rhodamine B(RhB)removal rate of 97%,much higher than the combination of electrocatalysis(EC,12%)and photocatalysis(PC,50%),confirming the synergy of photoelectrocatalysis.展开更多
基金supported in part by the National Natural Science Foundation of China under Grant 61873277in part by the Natural Science Basic Research Plan in Shaanxi Province of China underGrant 2020JQ-758in part by the Chinese Postdoctoral Science Foundation under Grant 2020M673446.
文摘In the video captioning methods based on an encoder-decoder,limited visual features are extracted by an encoder,and a natural sentence of the video content is generated using a decoder.However,this kind ofmethod is dependent on a single video input source and few visual labels,and there is a problem with semantic alignment between video contents and generated natural sentences,which are not suitable for accurately comprehending and describing the video contents.To address this issue,this paper proposes a video captioning method by semantic topic-guided generation.First,a 3D convolutional neural network is utilized to extract the spatiotemporal features of videos during the encoding.Then,the semantic topics of video data are extracted using the visual labels retrieved from similar video data.In the decoding,a decoder is constructed by combining a novel Enhance-TopK sampling algorithm with a Generative Pre-trained Transformer-2 deep neural network,which decreases the influence of“deviation”in the semantic mapping process between videos and texts by jointly decoding a baseline and semantic topics of video contents.During this process,the designed Enhance-TopK sampling algorithm can alleviate a long-tail problem by dynamically adjusting the probability distribution of the predicted words.Finally,the experiments are conducted on two publicly used Microsoft Research Video Description andMicrosoft Research-Video to Text datasets.The experimental results demonstrate that the proposed method outperforms several state-of-art approaches.Specifically,the performance indicators Bilingual Evaluation Understudy,Metric for Evaluation of Translation with Explicit Ordering,Recall Oriented Understudy for Gisting Evaluation-longest common subsequence,and Consensus-based Image Description Evaluation of the proposed method are improved by 1.2%,0.1%,0.3%,and 2.4% on the Microsoft Research Video Description dataset,and 0.1%,1.0%,0.1%,and 2.8% on the Microsoft Research-Video to Text dataset,respectively,compared with the existing video captioning methods.As a result,the proposed method can generate video captioning that is more closely aligned with human natural language expression habits.
文摘Why do taxonomists matter? The work of taxonomists is often understated if not completely misunderstood. Without taxonomists, organisms cannot be accurately identified, neither can these organisms be given universally accepted names, and reliably positioned in the phylogenetic tree of life. Thanks to the work of taxonomists over the last 269 years since Carl Linnaeus established the binomial system, we can now measure the health and wealth of our biodiversity in a refined, science-based inventory prescribed by stringent nomenclatural rules.
基金mainly supported by National Science Fund for Distinguished Young Scholars (31625024)a grant (2008DP173354) from the Key Laboratory of the Zoological Systematics and Evolution of the Chinese Academy of Sciences。
文摘Taxonomy plays an important role in understanding the origin, evolution, and ecological functionality of biodiversity. There are large number of unknown species yet to be described by taxonomists, which together with their ecosystem services cannot be effectively protected prior to description. Despite this, taxonomy has been increasingly underrated insufficient funds and permanent positions to retain young talents. Further, the impact factordriven evaluation systems in China exacerbate this downward trend, so alternative evaluation metrics are urgently necessary. When the current generation of outstanding taxonomists retires,there will be too few remaining taxonomists left to train the next generation. In light of these challenges, all co-authors worked together on this paper to analyze the current situation of taxonomy and put out a joint call for immediate actions to advance taxonomy in China.
基金supported by the National Natural Science Foundation of China(83418083)the Natural Science Foundation of Fujian Province(2019J01230).
文摘Modification is one of the most important and effective methods to improve the photoelectrocatalytic(PEC)performance of ZnO.In this paper,the Ru_(x)Zn_(1-x)O/Ti electrodes were prepared by thermal decomposition method and the effect of Ru content on those electrodes' electronic structure was analyzed through the first-principles calculation.Various tests were also performed to observe the microstructures and PEC performance.The results showed that as the Ru^(4+) transferred into ZnO lattice and replaced a number of Zn^(2+),the conduction band of ZnO moved downward and the valence band went upward.The number of photogenerated electron-hole pairs increased as the impurity levels appeared in the band gap.In addition,ZnO nanorods exhibited a smaller grain size and a rougher surface under the effect of Ru.Meanwhile,the RuO_(2) nanoparticles on the surface of ZnO nanorods acted as the electron-transfer channel,helping electrons transfer to the counter electrode and delaying the recombination of the electron-hole pairs.Specifically,the Ru_(x)Zn_(1-x)O/Ti electrodes with 9.375 mol% Ru exhibited the best PEC performance with a rhodamine B(RhB)removal rate of 97%,much higher than the combination of electrocatalysis(EC,12%)and photocatalysis(PC,50%),confirming the synergy of photoelectrocatalysis.