Search

The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link. To predict the concentrations of PM2.5 and PM10 in the 128 cities of China, the discrete grey prediction model with fractional order accumulation (DFGM(1,1)) was used to predict the annual average PM2.5 and PM10 concentrations from 2019 to 2023. The result is as following: the annual average PM2.5 concentrations of Xi'an, Xuzhou, Ordos, Jingmen, Meizhou, Huizhou, Panzhihua, Kunming, Jixi and Yichun are increasing from 2019 to 2023. The annual average PM2.5 concentrations in the 118 other cities are decreasing from 2019 to 2023. While the annual average PM10 concentrations of Taiyuan, Ordos, Dongguan, Karamay, Foshan, Yichun, Qitaihe, Jinzhou and Heihe are increasing from 2019 to 2023. The annual average PM10 concentrations in the 119 other cities are decreasing from 2019 to 2023.

The microbial compositions, quality characteristics, and structural changes in fresh brown rice noodles (FBRN) during storage were investigated. Total plate count and mold and yeast counts increased while the pH decreased during storage. Metagenomic sequencing revealed that the microbial composition of FBRN changed throughout storage. A comprehensive investigation of the variation in lipid content demonstrated that hydrolytic rancidity was responsible for lipid deterioration. LF-NMR showed an increase in the proportion of bound water and a decrease in the proportion of free water in FBRN. Moreover, significant changes in edible qualities were observed. The cooking loss increased three-fold and noodles hardness reduced by approximately 23%. Further, the high initial aldehyde content of FBRN diminished almost completely, while that of alcohols and esters increased, leading to significant flavor deterioration. The correlation and factor analysis suggested that the TPC and MY counts could be used as key indicators of FBRN deterioration.

openaire: EC/H2020/780069/EU//MeMAD This year, the PicSOM group participated only in the Video to Text (VTT), Description Generation subtask. For our submitted runs we used either the MSR-VTT dataset only, or MS COCO and MSR-VTT jointly for training. We used LSTM recurrent neural networks to generate descriptions based on multi-modal features extracted from the videos. We submitted four runs: • PICSOM_1: uses ResNet features for initialising the LSTM generator, and object and scene-type detection features as persistent input to the generator which is trained on MS COCO + MSR-VTT, • PICSOM_2: uses ResNet and object detection features for initialisation, and is trained on MS COCO + MSR-VTT, this is the only run based on our new PyTorch codebase, • PICSOM_3: uses ResNet and video category features for initialisation, and trajectory and audio-visual embedding features for persistent features, trained on MSR-VTT only, • PICSOM_4: is the same as PICSOM_3 except that the audio-visual embedding feature has been replaced with audio class detection outputs. The most significant difference between our runs came from expanding the original MSR-VTT training dataset by including MS COCO, which contains images annotated with captions. Having a larger and more diverse training set seems to bring larger improvements to the performance measures than using more advanced features. This finding has been confirmed also by our post-submission experiments that we are still continuing.

In this paper, we presented a neural network system for DCASE 2018 task 2, general purpose audio tagging. We fine-tuned the Google AudioSet feature generation model with different settings for the given 41 classes on top of a fully connected layer with 100 units. Then we used the fine-tuned models to generate 128 dimensional features for each 0.960s audio. We tried different neural network structures including LSTM and multi-level attention models. In our experiments, the multi-level attention model has shown its superiority over others. Truncating the silence parts, repeating and splitting the audio into the fixed length, pitch shifting augmentation, and mixup techniques are all used in our experiments. The proposed system achieved a result with MAP@3 score at 0.936, which outperforms the baseline result of 0.704 and achieves top 8% in the public leaderboard.

Germination is considered to be an effective process for improving the nutritional quality and functionality of cereals. In this study, changes of nutritional ingredients, antinutritional components, chemical composition, and antioxidant activities of buckwheat seeds over 72 h of germination were investigated, and the reasons for these changes are discussed. With the prolonged germination time, the contents of crude protein, reducing sugar, total phenolics, total flavonoids, and condensed tannins increased significantly, while the levels of crude fat, phytic acid, and the activity of trypsin inhibitor decreased. Phenolic compounds, such as rutin, vitexin, isovitexin, orientin, isoorientin, chlorogenic acid, trans-3-hydroxycinnamic acid, and p-hydroxybenzoic acid increased significantly during the germination process, which may be due to the activation of phenylalanine ammonialyase. The improvement of flavonoids led to significant enhancement of the antioxidant activities of germinated buckwheat. Germinated buckwheat had better nutritional value and antioxidant activities than ungerminated buckwheat, and it represented an excellent natural source of flavonoids and phenolic compounds, especially rutin and C-glycosylflavones. Therefore, germinated buckwheat could be used as a promising functional food for health promotion.