Proteomic and transcriptomic analysis of visual long-term memory in Drosophila melanogaster

The fruit fly,Drosophila melanogaster,is able to discriminate visual landmarks and form visual long-term memory in a flight simulator.Studies focused on the molecular mechanism of long-term memory have shown that memory formation requires mRNA transcription and protein synthesis.However,little is known about the molecular mechanisms underlying the visual learning paradigm.The present study demonstrated that both spaced training procedure(STP)and consecutive training procedure(CTP)would induce long-term memory at 12 hour after training,and STP caused significantly higher 12-h memory scores compared with CTP.Labelfree quantification of liquid chromatography-tandem mass spectrometry(LC-MS/MS)and microarray were utilized to analyze proteomic and transcriptomic differences between the STP and CTP groups.Proteomic analysis revealed 30 up-regulated and 27 down-regulated proteins;Transcriptomic analysis revealed 145 up-regulated and 129 down-regulated genes.Among them,five candidate genes were verified by quantitative PCR,which revealed results similar to microarray.These results provide insight into the molecular components influencing visual long-term memory and facilitate further studies on the roles of identified genes in memory formation.

Cognitive Navigation for Intelligent Mobile Robots:A Learning-Based Approach With Topological Memory

Autonomous navigation for intelligent mobile robots has gained significant attention,with a focus on enabling robots to generate reliable policies based on maintenance of spatial memory.In this paper,we propose a learning-based visual navigation pipeline that uses topological maps as memory configurations.We introduce a unique online topology construction approach that fuses odometry pose estimation and perceptual similarity estimation.This tackles the issues of topological node redundancy and incorrect edge connections,which stem from the distribution gap between the spatial and perceptual domains.Furthermore,we propose a differentiable graph extraction structure,the topology multi-factor transformer(TMFT).This structure utilizes graph neural networks to integrate global memory and incorporates a multi-factor attention mechanism to underscore elements closely related to relevant target cues for policy generation.Results from photorealistic simulations on image-goal navigation tasks highlight the superior navigation performance of our proposed pipeline compared to existing memory structures.Comprehensive validation through behavior visualization,interpretability tests,and real-world deployment further underscore the adapt-ability and efficacy of our method.

Long-Term Electrophysiological and Behavioral Analysis on the Improvement of Visual Working Memory Load, Training Gains, and Transfer Benefits

Recent evidence demonstrates that with training, one can enhance visual working memory (VWM) capacity and attention over time in the near transfer tasks. Not only do these studies reveal the characteristics of VWM load and the influences of training, they may also provide insights into developing effective rehabilitation for patients with VWM deficiencies. However, few studies have investigated VWM over extended periods of time and evaluated transfer benefits on non-trained tasks. Here, we combined behavioral and electroencephalographical approaches to investigate VWM load, training gains, and transfer benefits. Our results reveal that VWM capacity is directly correlated to the difference of event-related potential waveforms. In particular, the “magic number 4” can be observed through the contralateral delay amplitude and the average capacity is 3.25-item over 15 participants. Furthermore, our findings indicate that VWM capacity can be improved through training;and after training exercises, participants from the training group are able to dramatically improve their performance. Likewise, the training effects on non-trained tasks can also be observed at the 12th week after training. Therefore, we conclude that participants can benefit from training gains, and augmented VWM capacity sustained over long periods of time on specific variety of tasks.

Evaluation of the long-term memory for thermosensation regulated by neuronal calcium sensor-1 in Caenorhabditis elegans

Objective To evaluate whether the thermotaxis tracking model is suitable for assessing long-term memory （LTM） in the nematode Caenorhabditis elegans. Methods Animals were trained at 20℃ overnight in presence of food. The percentage of animals performing isothermal tracking （IT） behavior was measured at different time intervals after the training. Results The percentage of animals performing IT behavior, the numbers of body bends inside and outside the training temperature, and the expression patterns of AFD and AIY neurons were similar to those in control animals at 36 and 48 h after training; whereas when extending to 60, 72, and 84 h, locomotory behavior defects were observed in the assayed animals, suggesting that this thermal tracking model is feasible for analyzing LTM at 36 and 48 h after training. Moreover, the percent-age of animals performing IT behavior was reduced at 18, 36, and 48 h after training in neuronal calcium sensor-1 gene （nsc-1） mutant animals compared with that in wild-type N2 animals. In addition, exposure to plumbum （Pb） significantly repressed the LTM at 18, 36, and 48 h after training in both wild-type N2 and ncs-1 mutant animals. Conclusion The thermotaxis tracking model is suitable for evaluating the LTM regulated by NCS-1, and can be employed for elucidating regulatory functions of specific genes or effects of stimuli on memory in C. elegans.

A Cognitive Memory-Augmented Network for Visual Anomaly Detection

With the rapid development of automated visual analysis,visual analysis systems have become a popular research topic in the field of computer vision and automated analysis.Visual analysis systems can assist humans to detect anomalous events(e.g.,fighting,walking alone on the grass,etc).In general,the existing methods for visual anomaly detection are usually based on an autoencoder architecture,i.e.,reconstructing the current frame or predicting the future frame.Then,the reconstruction error is adopted as the evaluation metric to identify whether an input is abnormal or not.The flaws of the existing methods are that abnormal samples can also be reconstructed well.In this paper,inspired by the human memory ability,we propose a novel deep neural network(DNN)based model termed cognitive memory-augmented network(CMAN)for the visual anomaly detection problem.The proposed CMAN model assumes that the visual analysis system imitates humans to remember normal samples and then distinguishes abnormal events from the collected videos.Specifically,in the proposed CMAN model,we introduce a memory module that is able to simulate the memory capacity of humans and a density estimation network that can learn the data distribution.The reconstruction errors and the novelty scores are used to distinguish abnormal events from videos.In addition,we develop a two-step scheme to train the proposed model so that the proposed memory module and the density estimation network can cooperate to improve performance.Comprehensive experiments evaluated on various popular benchmarks show the superiority and effectiveness of the proposed CMAN model for visual anomaly detection comparing with the state-of-the-arts methods.The implementation code of our CMAN method can be accessed at https://github.com/CMANcode/CMAN_pytorch.

The Screening of Genes Sensitive to Long-Term, Low-Level Microwave Exposure and Bioinformatic Analysis of Potential Correlations to Learning and Memory

Objective To gain a better understanding of gene expression changes in the brain following microwave exposure in mice. This study hopes to reveal mechanisms contributing to microwave-induced learning and memory dysfunction. Methods Mice were exposed to whole body 2100 MHz microwaves with specific absorption rates (SARs) of 0.45 W/kg, 1.8 W/kg, and 3.6 W/kg for 1 hour daily for 8 weeks. Differentially expressing genes in the brains were screened using high-density oligonucleotide arrays, with genes showing more significant differences further confirmed by RT-PCR. Results The gene chip results demonstrated that 41 genes (0.45 W/kg group), 29 genes (1.8 W/kg group), and 219 genes (3.6 W/kg group) were differentially expressed. GO analysis revealed that these differentially expressed genes were primarily involved in metabolic processes, cellular metabolic processes, regulation of biological processes, macromolecular metabolic processes, biosynthetic processes, cellular protein metabolic processes, transport, developmental processes, cellular component organization, etc. KEGG pathway analysis showed that these genes are mainly involved in pathways related to ribosome, Alzheimer's disease, Parkinson's disease, long-term potentiation, Huntington's disease, and Neurotrophin signaling. Construction of a protein interaction network identified several important regulatory genes including synbindin (sbdn), Crystallin (CryaB), PPP1CA, Ywhaq, Psap, Psmb1, Pcbp2, etc., which play important roles in the processes of learning and memory. Conclusion Long-term, low-level microwave exposure may inhibit learning and memory by affecting protein and energy metabolic processes and signaling pathways relating to neurological functions or diseases.

Analysis of influencing factors of visual working memory in young adult patients with schizophrenia

Background Patients with schizophrenia have general cognitive impairments, and the impairment of working memory is considered to be the basis of cognitive impairments. The research on visual working memory, one of the subcomponents, is getting more and more attention.However, the influencing factors which cause the deficits of visual working memory in patients with schizophrenia have not been clearly explained. To provide evidence for cognitive impairment interventions, the present study explored the factors influencing the deficits of patients' visual working memory.Aim The present study discussed the relevant factors influencing the visual working memory of patients with schizophrenia by measuring the accuracy of the visual working memory of patients with schizophrenia and healthy controls.Methods Colour-recall paradigm was employed to measure the accuracy of the visual working memory of 61 healthy controls and 61 patients who met the International Classification of Diseases, Tenth Revision diagnostic criteria for schizophrenia. The age range of subjects was18-50. Scale for the Assessment of Positive Symptoms(SAPS) and Scale for the Assessment of Negative Symptoms(SANS) were used to evaluate the patients' clinical symptoms.Results Compared with the healthy control group, the accuracy of visual working memory of patients with schizophrenia was significantly impaired(t=3.062,p=0.003). The accuracy of visual working memory of patients with schizophrenia was not related to age(r=0.023,p=0.860),the age of onset(r=-0.003,p=0.979),the duration of illness(r=-0.038, p=0.769),education level(r=-0.181, p=0.162),continuous working time before illness(r =-0.107, p=0.413) or the daily dose of antipsychotic drugs(r =0.062, p=0.635); however, it was positively related to the number of hospitalisations(r =0.471,p<0.001). The total score of Scale for the Assessment of Positive Symptoms(SAPS) was negatively related to the accuracy of visual working memory(r=-0.388, p=0.005), while the total score of Scale for the Assessment of Negative Symptoms(SANS)(r=0.416,p=0.001), the total score of diminished emotional expressiveness(r =0.352, p=0.005) and the total score of attention disorder(r =0.310, p=0.015) were positively related to the accuracy of visual working memory. Patients using a single drug and those using multiple drugs were compared with each other. They were not significantly different in age(t=0.010, p=0.992), the number of hospitalisations(t=0.656, p=0.514), the duration of illness(t=0.701, p=0.486), the total score of SANS(t=0.078,p=0.938), the total score of SAPS(t=1.815, p=0.079) and the daily dose of antipsychotic drugs(t=1.794, p=0.078).However, in order to explore whether single or combined drug use would affect the accuracy of visual working memory of patients with schizophrenia, the present study also compared these two groups' different SO values of the accuracy of visual working memory. The results showed that the accuracy of visual working memory of patients with schizophrenia with combined drug use was significantly better than that of patients with single drug use(t=2.515, p=0.015, independent sample t-test).Conclusion The present study indicates that the visual working memory of young adult patients with schizophrenia is impaired compared with the healthy people within the same age range. The impairment is more obvious in patients who have multiple hospitalisations and suffer from severe negative symptoms. The impairment in patients with more severe positive symptoms is not very obvious. Combined drug use is likely to alleviate the impairment.

Delayed improvements in visual memory task performance among chronic schizophrenia patients after high-frequency repetitive transcranial magnetic stimulation

BACKGROUND Cognitive impairments are core characteristics of schizophrenia,but are largely resistant to current treatments.Several recent studies have shown that highfrequency repetitive transcranial magnetic stimulation(rTMS)of the left dorsolateral prefrontal cortex(DLPFC)can reduce negative symptoms and improve certain cognitive deficits in schizophrenia patients.However,results are inconsistent across studies.AIM To examine if high-frequency rTMS of the DLPFC can improve visual memory deficits in patients with schizophrenia.METHODS Forty-seven chronic schizophrenia patients with severe negative symptoms on stable treatment regimens were randomly assigned to receive active rTMS to the DLPFC(n=25)or sham stimulation(n=22)on weekdays for four consecutive weeks.Patients performed the pattern recognition memory(PRM)task from the Cambridge Neuropsychological Test Automated Battery at baseline,at the end of rTMS treatment(week 4),and 4 wk after rTMS treatment(week 8).Clinical symptoms were also measured at these same time points using the Scale for the Assessment of Negative Symptoms(SANS)and the Positive and Negative Syndrome Scale(PANSS).RESULTS There were no significant differences in PRM performance metrics,SANS total score,SANS subscores,PANSS total score,and PANSS subscores between active and sham rTMS groups at the end of the 4-wk treatment period,but PRM performance metrics(percent correct and number correct)and changes in these metrics from baseline were significantly greater in the active rTMS group at week 8 compared to the sham group(all P<0.05).Active rTMS treatment also significantly reduced SANS score at week 8 compared to sham treatment.Moreover,the improvement in visual memory was correlated with the reduction in negative symptoms at week 8.In contrast,there were no between-group differences in PANSS total score and subscale scores at either week 4 or week 8(all P>0.05).CONCLUSION High-frequency transcranial magnetic stimulation improves visual memory and reduces negative symptoms in schizophrenia,but these effects are delayed,potentially due to the requirement for extensive neuroplastic changes within DLPFC networks.

Correlating learning and memory improvements to long-term potentiation in patients with brain injury

BACKGROUND： Brain injury patients often exhibit learning and memory functional deficits. Long-term potentiation （LTP） is a representative index for studying learning and memory cellular models; the LTP index correlates to neural plasticity. OBJECTIVE： This study was designed to investigate correlations of learning and memory functions to LTP in brain injury patients, and to summarize the research advancements in mechanisms underlying brain functional improvements after rehabilitation intervention. RETRIEVAL STRATEGY： Using the terms ＂brain injuries, rehabilitation, learning and memory, long-term potentiation＂, manuscripts that were published from 2000-2007 were retrieved from the PubMed database. At the same time, manuscripts published from 2000-2007 were also retrieved from the Database of Chinese Scientific and Technical Periodicals with the same terms in the Chinese language. A total of 64 manuscripts were obtained and primarily screened. Inclusion criteria： studies on learning and memory, as well as LTP in brain injury patients, and studies focused on the effects of rehabilitation intervention on the two indices; studies that were recently published or in high-impact journals. Exclusion criteria： repetitive studies. LITERATURE EVALUATION： The included manuscripts primarily focused on correlations between learning and memory and LTP, the effects of brain injury on learning and memory, as well as LTP, and the effects of rehabilitation intervention on learning and memory after brain injury. The included 39 manuscripts were clinical, basic experimental, or review studies. DATA SYNTHESIS： Learning and memory closely correlates to LTP. The neurobiological basis of learning and memory is central nervous system plasticity, which involves neural networks, neural circuits, and synaptic connections, in particular, synaptic plasticity. LTP is considered to be an ideal model for studying synaptic plasticity, and it is also a classic model for studying neural plasticity of learning and memory. Brain injury patients clinically present with various manifestations, such as paralysis and sensory disability, which closely correlate to injured regions. In addition, learning and memory abilities decrease in brain injury patients and LTP decreases following brain injury. Brain tissue injury will lead to brain functional deficits. Hippocampal LTP is very sensitive. Difficulties in LTP induction are apparent even prior to morphological changes in brain tissue. There are no specific treatments for learning and memory functional deficits following brain injury. At present, behavioral and compensative therapies are the typical forms of rehabilitation. These results indicate that rehabilitation promotes learning and memory functional recovery in brain injury patients by speeding up LTP formation in the hippocampal CA3 region. CONCLUSION： Rehabilitation intervention increases LTP formation in the hippocampal CA3 region and recovers learning and memory functions in brain injury patients.

Study of Visual Perception and Working Memory in Moroccan Adolescents Attending School

Visual Perception and working memory are essential neuropsychological skills for any learning. The objective of this research is to study these skills, by analyzing numerical plots, according to gender, age, level of study and the stream by evaluating perceptive and memory deficits in Moroccan adolescents attending school. Our sample contains 146 high school students, including 78 boys and 68 girls participated in this study and we used the numerical version of The Rey-Osterrieth complex figure test (ROCF-A) to assess visual perception and working memory. The results of this study showed that, on the one hand, more than 34% of the subjects examined show signs of perceptual deficit and more 21% show signs of memory deficit and, on the other hand, that the scores of the visual perception are significantly associated with age and that working memory scores are significantly associated with age, level of education, and stream.

The Effect of a Visual Memory Training Program on Chinese Handwriting Performance of Primary School Students with Dyslexia in Hong Kong

This study investigated the effect of a visual memory training program on Chinese handwriting performance among primary school students with dyslexia in Hong Kong. Eight students of Grade 2 to 3 who were diagnosed with dyslexia were recruited. All participants received six sessions of training, which composed of 30-minute computerized game-based visual memory training and 30-minute Chinese character segmentation training. Visual perceptual skills and Chinese handwriting performance were assessed before and after the training, as well as three weeks after training using the Test of Visual Perceptual Skills (3rd edition) (TVPS-3) and the Chinese Handwriting Analysis System (CHAS). In comparing the pre- and post-training results, paired t-tests revealed significant improvements in visual memory skills, as well as handwriting speed, pause time and pen pressure after the training. There was no significant improvement in handwriting accuracy or legibility. The improved visual memory and handwriting performance did not show a significant drop at the follow-up assessments. This study showed promising results on a structured program to improve the Chinese handwriting performance, mainly in speed, of primary school children. The improvements appeared to be well-sustained after the training program. There is a need to further study the long-term effect of the program through a randomized controlled trial study.

Analyses of fear memory in Arc/Arg3.1-deficient mice: intact short-term memory and impaired long-term and remote memory

Activity-regulated cytoskeleton-associated protein (Arc/Arg3.1) was originally identified in patients with seizures. It is densely distributed in the hip-pocampus and amygdala in particular. Because the expression of Arc/Arg3.1 is regulated by nerve in-puts, it is thought to be an immediate early gene. As shown both in vitro and in vivo, Arc/Arg3.1 is in-volved in synaptic consolidation and regulates some forms of learning and memory in rats and mice [1,2]. Furthermore, a recent study suggests that Arc/Arg3.1 may play a significant role in signal transmission via AMPA-type glutamate receptors [3-5]. Therefore, we conducted a detailed analysis of fear memory in Arc/Arg3.1-deficient mice. As previously reported, the knockout animals exhib-ited impaired fear memory in both contextual and cued test situations. Although Arc/Arg3.1-deficient mice showed almost the same performance as wild-type littermates 4 hr after a conditioning trial, their performance was impaired in the retention test after 24 hr or longer, either with or without reconsolidation. Immunohistochemical analyses showed an abnormal density of GluR1 in the hip-pocampus of Arc/Arg3.1-deficient mice;however, an application of AMPA potentiator did not improve memory performance in the mutant mice. Memory impairment in Arc/Arg3.1-deficient mice is so ro-bust that the mice provide a useful tool for devel-oping treatments for memory impairment.

IoT-Driven Optimal Lightweight RetinaNet-Based Object Detection for Visually Impaired People

Visual impairment is one of the major problems among people of all age groups across the globe.Visually Impaired Persons(VIPs)require help from others to carry out their day-to-day tasks.Since they experience several problems in their daily lives,technical intervention can help them resolve the challenges.In this background,an automatic object detection tool is the need of the hour to empower VIPs with safe navigation.The recent advances in the Internet of Things(IoT)and Deep Learning(DL)techniques make it possible.The current study proposes IoT-assisted Transient Search Optimization with a Lightweight RetinaNetbased object detection(TSOLWR-ODVIP)model to help VIPs.The primary aim of the presented TSOLWR-ODVIP technique is to identify different objects surrounding VIPs and to convey the information via audio message to them.For data acquisition,IoT devices are used in this study.Then,the Lightweight RetinaNet(LWR)model is applied to detect objects accurately.Next,the TSO algorithm is employed for fine-tuning the hyperparameters involved in the LWR model.Finally,the Long Short-Term Memory(LSTM)model is exploited for classifying objects.The performance of the proposed TSOLWR-ODVIP technique was evaluated using a set of objects,and the results were examined under distinct aspects.The comparison study outcomes confirmed that the TSOLWR-ODVIP model could effectually detect and classify the objects,enhancing the quality of life of VIPs.

The Function of Oxytocin in Memory —A General Review of Oxytocin’s Effect on Memory

Introduction: While oxytocin (OT) is widely recognized for its pivotal role in reproductive behavior and the formation of social bonds, there remains a significant gap in our understanding of its potential influence on learning and memory processes, encompassing both social and non-social aspects. Thus this paper serves as an attempt to investigate the comprehensive role of OT in Physiological, Cognitive, and Behavioral processes. Method: A comprehensive literature review was conducted to assemble evidence related to the influence of OT on learning and memory. Studies encompassing both social and non-social memory were incorporated into the analysis. Additionally, molecular mechanisms through which OT could potentially impact neuronal activity in the hippocampus and amygdala, consequently affecting learning and memory, were also investigated. Results: Our review reveals a spectrum of evidence that both supports and contradicts the theory that OT plays a significant role in social and non-social memory. While certain studies suggest a positive impact of OT on memory, others present findings that argue otherwise. However, multiple potential molecular mechanisms were discovered that may elucidate OT’s effects on learning and memory, particularly its potential to modulate neuronal activity in the hippocampus and amygdala. Conclusion: Despite the mixed evidence, OT might have a significant role in both social and non-social memory. Identified molecular mechanisms propose potential ways in which OT could influence learning and memory. The key role appears to be the modulation of neuronal activity in the hippocampus and amygdala by OT. Furthermore, it is plausible that OT’s function in memory is crucial for the social behaviors previously associated with it. Future research is necessitated to fully unravel the exact mechanisms and implications of OT’s role in learning and memory.

卒中后失语工作记忆的事件相关电位及时频特征

目的 探索卒中后失语(PSA)工作记忆的任务态脑电特征。方法 2020年9月至2021年2月,于北京中医药大学东直门医院招募PSA患者(PSA组)和健康受试者(HC组)各8例,采集脑电图和记忆量表数据,通过工作记忆任务态的脑电数据分析时域事件相关电位(ERP)和时频的脑电频段指标特点,与记忆量表中的各项目进行相关性分析。结果 最终纳入10例,每组5例。ERP分析发现前额区诱发N1和P2成分,顶枕区诱发P300成分。与HC组相比,PSA组在出现条件下,中央前额区N1和P2激活增强,右侧顶枕区P300活动降低(|t|> 2.193, P <0.05)。与HC组相比,PSA组在未出现条件下右侧前额区、中央顶枕区θ频段能量降低,左侧顶枕区α1频段能量降低,左侧中央区γ频段能量增加(t> 2.398, P <0.05)。相关性分析显示,γ频段的能量与听觉词语学习测验中的即刻回忆(r=0.914, P=0.030)、正确再认(r=0.931, P=0.022)和数字广度测验中的倒背(r=0.924, P=0.025)、顺背(r=0.889, P=0.044)呈强正相关。结论 视觉工作记忆任务可调动PSA患者记忆相关脑区代偿活动,可作为PSA工作记忆相关研究的评价客观指征。语言损伤与工作记忆之间存在密切联系。

基于多源信息融合和WOA-CNN-LSTM的外脚手架隐患分类预警研究

面对施工现场外脚手架隐患信息的多样性,传统的基于传感器监测的单一信号预警研究存在容错力不佳、含有信息有限等问题。针对施工现场外脚手架“图像+监测”数据,提出一种基于数据层和特征层信息融合的脚手架隐患分类预警方法。首先,利用Revit三维建模软件建立外脚手架实体模型,对不同初始隐患下的外脚手架进行有限元分析,划分隐患预警等级;其次,利用无迹卡尔曼滤波算法(Unscented Kalman Filter,UKF)及卷积长短时记忆网络(Convolutional Neural Network-Long Short Term Memory Network,CNN-LSTM)实现脚手架同类信息数据层融合及异类信息特征层融合;最后,通过实时收集西安市某在建项目落地式双排扣件式钢管脚手架隐患信息,对其进行分类预警,并使用鲸鱼优化算法(Whale Optimization Algorithm,WOA)对CNN-LSTM网络进行参数优化,发现隐藏节点个数为30、学习率为0.0072、正则化系数为1×10^(-4)时分类效果最佳,优化后预警精度达到了91.4526%。通过可视化WOA-CNN-LSTM、CNN-LSTM、CNN-SVM(Support Vector Machine,支持向量机)及CNN-GRU(Gate Recurrent Unit,门控循环单元)分类预警结果,证实了优化后的CNN-LSTM网络在脚手架分类预警方面的优越性。

从汉简看汉代奏议书写形态的法定性与视觉修辞

奏议起源于先秦的朝堂议政,是臣僚向君王谢恩、陈请、议政、进谏的文书。在汉代“以文御国”的背景下,出于行政需要,文质兼备的奏议章表应运勃兴。奏议文书往往“因事而发”,蕴含着非凡的政治智慧,具有政治实用性;其内容、格式、书写等具有固定的套语格式和鲜明的制度化特征。其文本对特定称谓和程序化语言的运用,表明其书写形态的法定性。传世汉代奏议多出自史书载录,大多曾经史家“裁剪”,“过滤”了其书写形态运用视觉修辞的诸多“细节”,有碍于一窥奏议之“全豹”。借助西北汉简实物,可直观奏议文本特定称谓与程序化语言及其书写借助字体大小、隶草变化、笔划夸饰等书写视觉效果变化,突出奏议文本物化形态的视觉修辞策略,借以强调文本书写与礼制的内在联系。

长视频的超级帧切割视觉内容解释方法

针对现有基于编码解码的视频描述方法存在的对视频较长、在视频场景切换频繁情况下视觉特征提取能力不足或关键性片段捕获能力不足等视频描述不佳的问题,提出一种基于超级帧切割长视频的视频字幕方法。首先,提出超级帧提取算法,计算关键视频时间占比率以满足视频浏览时长限制,缩短视频检索时间。然后,构建两层筛选模型以自适应提取超级帧,过滤冗余关键帧,执行多场景语义描述。将保留的关键帧嵌入周围帧,利用深层网络模型以及小卷积核池化采样域获取更多的视频特征,克服了经典视频标题方法不能直接用于处理长视频的困难。最后,通过用长短时记忆模型代替循环神经网络解码生成视频标题,给出视频内容的分段解释信息。在YouTube数据集视频、合成视频和监控长视频上进行测试,采用多种机器翻译评价指标评估了该方法的性能,均获得了不同程度的提升。实验结果表明,该方法在应对视频场景切换频繁、视频较长等挑战时,能够获得较好的片段描述。