Deep learning with convolutional neural networks(CNNs)has achieved great success in the classification of various plant diseases.However,a limited number of studies have elucidated the process of inference,leaving it ...Deep learning with convolutional neural networks(CNNs)has achieved great success in the classification of various plant diseases.However,a limited number of studies have elucidated the process of inference,leaving it as an untouchable black box.Revealing the CNN to extract the learned feature as an interpretable form not only ensures its reliability but also enables the validation of the model authenticity and the training dataset by human intervention.In this study,a variety of neuron-wise and layer-wise visualization methods were applied using a CNN,trained with a publicly available plant disease image dataset.We showed that neural networks can capture the colors and textures of lesions specific to respective diseases upon diagnosis,which resembles human decision-making.While several visualizationmethods were used as they are,others had to be optimized to target a specific layer that fully captures the features to generate consequential outputs.Moreover,by interpreting the generated attention maps,we identified several layers that were not contributing to inference and removed such layers inside the network,decreasing the number of parameters by 75%without affecting the classification accuracy.The results provide an impetus for the CNN black box users in the field of plant science to better understand the diagnosis process and lead to further efficient use of deep learning for plant disease diagnosis.展开更多
With the development of intelligent agents pursuing humanisation,artificial intelligence must consider emotion,the most basic spiritual need in human interaction.Traditional emotional dialogue systems usually use an e...With the development of intelligent agents pursuing humanisation,artificial intelligence must consider emotion,the most basic spiritual need in human interaction.Traditional emotional dialogue systems usually use an external emotional dictionary to select appropriate emotional words to add to the response or concatenate emotional tags and semantic features in the decoding step to generate appropriate responses.However,selecting emotional words from a fixed emotional dictionary may result in loss of the diversity and consistency of the response.We propose a semantic and emotion-based dual latent variable generation model(Dual-LVG)for dialogue systems,which is able to generate appropriate emotional responses without an emotional dictionary.Different from previous work,the conditional variational autoencoder(CVAE)adopts the standard transformer structure.Then,Dual-LVG regularises the CVAE latent space by introducing a dual latent space of semantics and emotion.The content diversity and emotional accuracy of the generated responses are improved by learning emotion and semantic features respectively.Moreover,the average attention mechanism is adopted to better extract semantic features at the sequence level,and the semi-supervised attention mechanism is used in the decoding step to strengthen the fusion of emotional features of the model.Experimental results show that Dual-LVG can successfully achieve the effect of generating different content by controlling emotional factors.展开更多
In order to further understand the mechanism of material volume change in the drying process,numerical simulations(considering or neglecting shrinkage)of heat and mass transfer during convective drying of carrot slice...In order to further understand the mechanism of material volume change in the drying process,numerical simulations(considering or neglecting shrinkage)of heat and mass transfer during convective drying of carrot slices under constant and controlled temperature and relative humidity were carried out.Simulated results were validated with experimental data.The results of the simulation show that the Quadratic model fitted well to the moisture ratio and the material temperature data trend with average relative errors of 5.9%and 8.1%,respectively.Additionally,the results of the simulation considering shrinkage show that the moisture and temperature distributions during drying are closer to the experimental data than the results of the simulation disregarding shrinkage.The material moisture content was significantly related to the shrinkage of dried tissue.Temperature and relative humidity significantly affected the volume shrinkage of carrot slices.The volume shrinkage increased with the rising of the constant temperature and the decline of relative humidity.This model can be used to provide more information on the dynamics of heat and mass transfer during drying and can also be adapted to other products and dryers devices.展开更多
Image matting is to estimate the opacity of foreground objects from an image. A few deep learning based methods have been proposed for image matting and perform well in capturing spatially close information. However, ...Image matting is to estimate the opacity of foreground objects from an image. A few deep learning based methods have been proposed for image matting and perform well in capturing spatially close information. However, these methods fail to capture global contextual information, which has been proved essential in improving matting performance. This is because a matting image may be up to several megapixels, which is too big for a learning-based network to capture global contextual information due to the limit size of a receptive field. Although uniformly downsampling the matting image can alleviate this problem, it may result in the degradation of matting performance. To solve this problem, we introduce a natural image matting with the attended global context method to extract global contextual information from the whole image, and to condense them into a suitable size for learning-based network. Specifically, we first leverage a deformable sampling layer to obtain condensed foreground and background attended images respectively. Then, we utilize a contextual attention layer to extract information related to unknown regions from condensed foreground and background images generated by a deformable sampling layer. Besides, our network predicts a background as well as the alpha matte to obtain more purified foreground, which contributes to better qualitative performance in composition. Comprehensive experiments show that our method achieves competitive performance on both Composition-1k and the alphamatting.com benchmark quantitatively and qualitatively.展开更多
This research explored the application of pulsed vacuum technology on the drying(PVD)of pineapple slices.Influences of drying temperature and pulsed vacuum ratio(PVR)on drying characteristics and pineapple quality(col...This research explored the application of pulsed vacuum technology on the drying(PVD)of pineapple slices.Influences of drying temperature and pulsed vacuum ratio(PVR)on drying characteristics and pineapple quality(color,rehydration characteristics,microstructure,and texture)were analyzed.As expected,increasing the drying temperature resulted in a higher drying rate and effective moisture diffusivity.The optimal PVR of 5:5 was beneficial in accelerating the drying rate of pineapple slices and the corresponding effective moisture diffusion coefficient(8.9601×10^(-10))was higher than other PVR conditions based on material center temperature.The material temperature increased during the normal pressure period and decreased rapidly when the pressure dropped to the vacuum condition,which indirectly reflected the moisture transfer that occurred during the vacuum holding period,while moisture diffusion happened during the atmospheric pressure holding period.The optimal pulsed vacuum drying process(PVR of 5:5)could expand air and water vapor and create a looser structure so as to obtain better rehydration performance(rehydration ratio(RR)was 5.43).High drying temperature led to the decrease of L^(*)value,the increase ofΔE value,and even the formation of surface scorch at 80℃.At the same drying temperature,the color quality depended on the drying time,and the color difference increased with the extension of the drying time.The chewiness and hardness of pineapple slices dried by PVD were significantly higher than those of fresh samples,which was conducive to the chewing taste.展开更多
In this paper, we present a video coding scheme which applies the technique of visual saliency computation to adjust image fidelity before compression. To extract visually salient features, we construct a spatio-tempo...In this paper, we present a video coding scheme which applies the technique of visual saliency computation to adjust image fidelity before compression. To extract visually salient features, we construct a spatio-temporal saliency map by analyzing the video using a combined bottom-up and top-down visual saliency model. We then use an extended bilateral filter, in which the local intensity and spatial scales are adjusted according to visual saliency, to adaptively alter the image fidelity. Our implementation is based on the H.264 video encoder JM12.0. Besides evaluating our scheme with the H.264 reference software, we also compare it to a more traditional foreground-background segmentation-based method and a foveation-based approach which employs Gaussian blurring. Our results show that the proposed algorithm can improve the compression ratio significantly while effectively preserving perceptual visual quality.展开更多
基金This research was supported by Japan Science and Tech-nology Agency (JST) PRESTO[Grants nos.JPMJPR17O5(Yosuke'Toda) and JPMJPR17O3(Fumio Okura)].
文摘Deep learning with convolutional neural networks(CNNs)has achieved great success in the classification of various plant diseases.However,a limited number of studies have elucidated the process of inference,leaving it as an untouchable black box.Revealing the CNN to extract the learned feature as an interpretable form not only ensures its reliability but also enables the validation of the model authenticity and the training dataset by human intervention.In this study,a variety of neuron-wise and layer-wise visualization methods were applied using a CNN,trained with a publicly available plant disease image dataset.We showed that neural networks can capture the colors and textures of lesions specific to respective diseases upon diagnosis,which resembles human decision-making.While several visualizationmethods were used as they are,others had to be optimized to target a specific layer that fully captures the features to generate consequential outputs.Moreover,by interpreting the generated attention maps,we identified several layers that were not contributing to inference and removed such layers inside the network,decreasing the number of parameters by 75%without affecting the classification accuracy.The results provide an impetus for the CNN black box users in the field of plant science to better understand the diagnosis process and lead to further efficient use of deep learning for plant disease diagnosis.
基金Fundamental Research Funds for the Central Universities of China,Grant/Award Number:CUC220B009National Natural Science Foundation of China,Grant/Award Numbers:62207029,62271454,72274182。
文摘With the development of intelligent agents pursuing humanisation,artificial intelligence must consider emotion,the most basic spiritual need in human interaction.Traditional emotional dialogue systems usually use an external emotional dictionary to select appropriate emotional words to add to the response or concatenate emotional tags and semantic features in the decoding step to generate appropriate responses.However,selecting emotional words from a fixed emotional dictionary may result in loss of the diversity and consistency of the response.We propose a semantic and emotion-based dual latent variable generation model(Dual-LVG)for dialogue systems,which is able to generate appropriate emotional responses without an emotional dictionary.Different from previous work,the conditional variational autoencoder(CVAE)adopts the standard transformer structure.Then,Dual-LVG regularises the CVAE latent space by introducing a dual latent space of semantics and emotion.The content diversity and emotional accuracy of the generated responses are improved by learning emotion and semantic features respectively.Moreover,the average attention mechanism is adopted to better extract semantic features at the sequence level,and the semi-supervised attention mechanism is used in the decoding step to strengthen the fusion of emotional features of the model.Experimental results show that Dual-LVG can successfully achieve the effect of generating different content by controlling emotional factors.
基金supported by Earmarked Fund for China Agriculture Research System(CARS-21).
文摘In order to further understand the mechanism of material volume change in the drying process,numerical simulations(considering or neglecting shrinkage)of heat and mass transfer during convective drying of carrot slices under constant and controlled temperature and relative humidity were carried out.Simulated results were validated with experimental data.The results of the simulation show that the Quadratic model fitted well to the moisture ratio and the material temperature data trend with average relative errors of 5.9%and 8.1%,respectively.Additionally,the results of the simulation considering shrinkage show that the moisture and temperature distributions during drying are closer to the experimental data than the results of the simulation disregarding shrinkage.The material moisture content was significantly related to the shrinkage of dried tissue.Temperature and relative humidity significantly affected the volume shrinkage of carrot slices.The volume shrinkage increased with the rising of the constant temperature and the decline of relative humidity.This model can be used to provide more information on the dynamics of heat and mass transfer during drying and can also be adapted to other products and dryers devices.
基金supported by the National Natural Science Foundation of China under Grant No.62076162the Shanghai Municipal Science and Technology Major Project under Grant Nos.2021SHZDZX0102 and 20511100300.
文摘Image matting is to estimate the opacity of foreground objects from an image. A few deep learning based methods have been proposed for image matting and perform well in capturing spatially close information. However, these methods fail to capture global contextual information, which has been proved essential in improving matting performance. This is because a matting image may be up to several megapixels, which is too big for a learning-based network to capture global contextual information due to the limit size of a receptive field. Although uniformly downsampling the matting image can alleviate this problem, it may result in the degradation of matting performance. To solve this problem, we introduce a natural image matting with the attended global context method to extract global contextual information from the whole image, and to condense them into a suitable size for learning-based network. Specifically, we first leverage a deformable sampling layer to obtain condensed foreground and background attended images respectively. Then, we utilize a contextual attention layer to extract information related to unknown regions from condensed foreground and background images generated by a deformable sampling layer. Besides, our network predicts a background as well as the alpha matte to obtain more purified foreground, which contributes to better qualitative performance in composition. Comprehensive experiments show that our method achieves competitive performance on both Composition-1k and the alphamatting.com benchmark quantitatively and qualitatively.
基金supported in part by the Science and Technology Program of Hebei(Grant No.203777119D,19227210D)in part by the Scientific Research Projects of Universities in Hebei Province(Grant No.ZD2021056).
文摘This research explored the application of pulsed vacuum technology on the drying(PVD)of pineapple slices.Influences of drying temperature and pulsed vacuum ratio(PVR)on drying characteristics and pineapple quality(color,rehydration characteristics,microstructure,and texture)were analyzed.As expected,increasing the drying temperature resulted in a higher drying rate and effective moisture diffusivity.The optimal PVR of 5:5 was beneficial in accelerating the drying rate of pineapple slices and the corresponding effective moisture diffusion coefficient(8.9601×10^(-10))was higher than other PVR conditions based on material center temperature.The material temperature increased during the normal pressure period and decreased rapidly when the pressure dropped to the vacuum condition,which indirectly reflected the moisture transfer that occurred during the vacuum holding period,while moisture diffusion happened during the atmospheric pressure holding period.The optimal pulsed vacuum drying process(PVR of 5:5)could expand air and water vapor and create a looser structure so as to obtain better rehydration performance(rehydration ratio(RR)was 5.43).High drying temperature led to the decrease of L^(*)value,the increase ofΔE value,and even the formation of surface scorch at 80℃.At the same drying temperature,the color quality depended on the drying time,and the color difference increased with the extension of the drying time.The chewiness and hardness of pineapple slices dried by PVD were significantly higher than those of fresh samples,which was conducive to the chewing taste.
基金supported partially by the National High-Tech Research and Development 863 Program of China under Grant No. 2009AA01Z330the National Natural Science Foundation of China under Grant Nos.61033012 and 60970100
文摘In this paper, we present a video coding scheme which applies the technique of visual saliency computation to adjust image fidelity before compression. To extract visually salient features, we construct a spatio-temporal saliency map by analyzing the video using a combined bottom-up and top-down visual saliency model. We then use an extended bilateral filter, in which the local intensity and spatial scales are adjusted according to visual saliency, to adaptively alter the image fidelity. Our implementation is based on the H.264 video encoder JM12.0. Besides evaluating our scheme with the H.264 reference software, we also compare it to a more traditional foreground-background segmentation-based method and a foveation-based approach which employs Gaussian blurring. Our results show that the proposed algorithm can improve the compression ratio significantly while effectively preserving perceptual visual quality.
