Subjects: Psychology >> Social Psychology submitted time 2023-03-28 Cooperative journals: 《心理科学进展》
Abstract： Studies regarding the multicomponent model of working memory mainly focus on the storage of featural properties, spatiotemporal properties and verbal information of objects, as well as the binding of these information (e.g., Allen et al., 2015; Fellman et al., 2017; Logie, 1995; Son et al., 2020; Zhao et al., 2019). The storage of limb movement information has not been explored. Limb movements are one of the important ways individuals interact with their environment. Exploring the storage of limb movement information is helpful to deeply clarify the storage modes of various types of information, as well as understand how different types of information transcoded and interacted with each other. Smyth et al. (1988) proposed two types of limb movements, i.e., movement pattern (including a gesture or movement to be imitated, such as an arabesque in ballet) and movement to positions in space (such as picking up a pen) according to the different goals of movements. The goal of movement pattern is the body pattern, whereas achieving a spatial target is the goal of movement to positions in space. In other words, movement pattern refers to kinesthetic or motor coding in imitation; movement to positions in space refers to the use of movement in visuo-spatial processing. In the field of perception and working memory for limb movement, previous studies did not regard the two types of limb movements as a whole. On the contrary, they usually explored the storage of information of movement to positions in space and body movement patterns information respectively or even compared them in one study. Based on this, the current study reviewed and compared the storage mechanism of these two types of limb movement information. Studies on movement to positions in space have revealed that the working memory task of movement to positions affects the encoding of spatial working memory, but it is separated from visual working memory and verbal working memory. In addition, information of movement to positions in space shared brain area (the superior parietal lobule) with spatial information of the object rather than with verbal information and information of the object’s featural properties; information of movement to positions in space activates unique brain areas (the contralateral motor cortex, the primary motor cortex, the ventral supplementary motor area, the left supramotor cortical areas and the primary motor cortex, etc.) that are independent of the other three kinds of information. Researches on body movement patterns have revealed that working memory for body movement patterns and verbal working memory are separated. In addition, the storage of body movement patterns only activates the brain regions that store spatial information of the object, rather than the brain regions that store information of the object’s featural properties and verbal information. More importantly, only the storage of body movement patterns activates the movement-related cortex (the middle temporal). Therefore, the storage of two kinds of limb movement information is independent of the phonological loop and the visual subsystem in the visuospatial sketchpad and needs the participation of the spatial subsystem in the visuospatial sketchpad; movement to positions in space and body movement patterns activate different movement-related cortexes that are independent of the phonological loop, the visual subsystem and the spatial subsystem in the visuospatial sketchpad. These results show that the existing multicomponent model of working memory cannot fully explain the storage of limb movement information. It is implied that there is a “limb movement system” in the working memory system that is specific to limb movement information, belongs to visuospatial sketchpad and coexists with the visual subsystem and spatial subsystem. The brain areas activated in the “limb movement system” vary with different kinds of limb movements.
Subjects: Psychology >> Social Psychology submitted time 2023-03-28 Cooperative journals: 《心理科学进展》
Abstract： The approximate number system plays an important role in the development of individual mathematical abilities, and the relationship between the two factors is affected by age. Mainly, as age increases, the degree of correlation gradually weakens, and the mechanisms change from cardinal knowledge mediation to the joint effect of multiple intermediary variables. Future research should use a more rigorous experimental design and multiple research methods to investigate the development trend, causal direction, key turning points and the underlying mechanisms of the relationship between the approximate number system and different mathematical abilities of children of all ages to better understand the role of the approximate number system in the development of individual mathematical abilities.
Subjects: Biology >> Botany >> Applied botany submitted time 2021-12-19 Cooperative journals: 《广西植物》
Abstract： In order to explore the response mechanism of AP2/ERF gene family in the water stress of Olea europaea, this study performed transcriptome sequencing on the roots and leaves of two cultivars 'Frantoio' and 'TYZ-1' that were under drought and flooding stress. And based on the whole genome data, the protein physicochemical properties, gene structure and system evolution of AP2/ERF transcription factor in O. europaea were analyzed. At the same time, the difference in gene expression of AP2/ERF transcription factor related to water stress in the two O. europaea cultivars was analyzed by transcriptome sequencing data and verified by fluorescence quantitative PCR. The results were as follows: (1) A total of 110 AP2/ERF gene family members were identified in O. europaea. The amino acid size of the 110 proteins is 173-717bp, there is no signal peptide and it is a non-secreted protein. The phylogenetic tree was constructed between O. europaea AP2/ERF and model plant Arabidopsis AP2/ERF protein. It was found that O. europaea AP2/ERF protein was divided into four categories: AP2, RAV, ERF and Solosist. Among them, ERF is divided into two subtypes: ERF and DREB. ERF includes six subtypes of ERF B1 to ERF B6, and DREB includes six subtypes of DREB A1 to DREB A6, which is consistent with the classification of the model plant Arabidopsis AP2/ERF. Each subfamily contains AP2/ERF proteins of O. europaea and Arabidopsis at the same time, indicating that the AP2/ERF family of Arabidopsis and O. europaea are similar in evolution. (2) The analysis of gene structure and conserved elements found that the proteins of the same subfamily of O. europaea AP2/ERF have the same gene structure and conserved elements. Combining gene expression with genes with known water regulation functions in the evolutionary tree, it is preliminarily speculated that OeAP2-75, OeAP2-97, OeAP2-101, OeAP2-23, OeAP2-13 are closely related to the water regulation of O. europaea, OeAP2-13, OeAP2-28, OeAP2-104, OeAP2-75, OeAP2-80, OeAP2-50 have different expression levels in the two varieties. It is speculated that this may be the reason for the different resistance of 'Frantoio' and 'TYZ-1'. The RT-qPCR technique was used to detect the expression changes of O. europaea AP2/ERF gene under different stresses. The results showed that OeAP2-101, OeAP2-28, and OeAP2-42 were significantly up-regulated by water stress, which was consistent with the results of transcriptome analysis. The results of this study can lay a foundation for the research on the stress resistance expression and gene function of the AP2/ERF family genes of O. europaea, and provide a method and theoretical basis for the selection of drought-resistant and flood-tolerant rootstock varieties of O. europaea.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2019-01-28 Cooperative journals: 《计算机应用研究》
Abstract： The existing mainstream methods use the pre-trained convolutional neural networks to extract image features and usually have the following limitations: a）Only using a single layer of pre-trained features to represent image; b）Inconsistency between the pre-trained task and the actual research task. These limitations result in that the existing methods of image-text matching cannot make full use of image features and is easily influenced by the noises. To solve the above limitations, this paper used multi-layer features from a pre-trained network and proposed a fusion algorithm of multi-level image features accordingly. Under the guidance of the image-text matching objective function, the proposed algorithm fused the multi-level pre-trained image features and reduced their dimensionality using a multi-layer perceptron to generate fusion features. It is able to make full use of pre-trained features and successfully reduce the influences of noises. The experimental results show that the proposed fusion algorithm makes better use of pre-trained image features and outperforms the methods using single-level features in the image-text matching task.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-11-29 Cooperative journals: 《计算机应用研究》
Abstract： The method based on recurrent neural network combined with syntactic structure is widely used in relation classification, and the neural network is used to automatically acquire features and realize relation classification. However, the existing methods are mainly based on a single specific syntactic structure model, and the model of a specific syntactic structure cannot be transferred to other types of syntactic structures. Aiming at this problem, a hierarchical recurrent neural network model with multi-syntactic structure is proposed. The hierarchical recurrent neural network is divided into two layers for network construction. Firstly, entity pre-training is performed in the sequence layer. The Bi-LSTM-CRF fusion Attention mechanism is used to improve the model's attention to the entity information on the text sequence, thereby obtaining more accurate. The more accurate entity feature information promotes better classification in the relation layer stage. Secondly, in the relation layer, the Bi-Tree-LSTM is nested above the sequence layer, and the hidden state and entity feature information of the sequence layer is passed into the relation layer, then three different syntax structures are weighted learned using the shared parameters and classify the semantic relation finally. The experimental results show that the model has a marco-F1 value of 85.9% on the SemEval-2010 Task8 corpus, and further improves the robustness of the model.
Hits
Hits
Downloads
Comment