检索结果-内蒙古大学图书馆

Music emotion prediction based on multidimensional feature fusion

Music emotion prediction based on multidimensional feature f...

Information technology,Big Data and Artificial Intelligence (ICIBA), International Conference on IEEE International Conference on

作者： Xuanye Zhou Zhibin Su Hui Ren School of Information and Communication Engineering Communication University of China Beijing China Key Laboratory of Acoustic Visual Technology and Intelligent Control System Ministry of Culture and Tourism Beijing China

Audience sentiment analysis is a popular research task in Music Information Retrieval (MIR). In this work, we aim to lean the prediction model of fine-grained music emotion for intelligent editing and retrieval. To satisfy the properties of time-series of music, the RNN-based model of BILSTM was selected as the bone. In particular, the traditional features and deep features extracted from the CNN architecture of spectrogram was fused with attention mechanism to enhance the recognition ability for different emotions. The ablation experiment has proved the necessary of the multi-dimensional features. Finally, the contrasting experiments on recent algorithms demonstrate that our proposed method for emotion prediction is effective, which is also suitable for small data sets.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Boosting Unsupervised Monocular Depth Estimation with Auxiliary Semantic Information

引用

China Communications 2021年第6期18卷 228-243页

作者： Hui Ren Nan Gao Jia Li State Key Laboratory of Media Convergence and Communication、Key Laboratory of Acoustic Visual Technology and Intelligent Control System Ministry of Culture and Tourism(Communication University of China)、Beijing Key Laboratory of Modern Entertainment Technology(Communication University of China)、School of Information and Communication EngineeringCommunication University of China.No.1Dingfuzhuang StreetChaoyang DistrictBeijing100024China

Learning-based multi-task models have been widely used in various scene understanding tasks,and complement each other,i.e.,they allow us to consider prior semantic information to better infer *** boost the unsupervised monocular depth estimation using semantic segmentation as an auxiliary *** address the lack of cross-domain datasets and catastrophic forgetting problems encountered in multi-task training,we utilize existing methodology to obtain redundant segmentation maps to build our cross-domain dataset,which not only provides a new way to conduct multi-task training,but also helps us to evaluate results compared with those of other *** addition,in order to comprehensively use the extracted features of the two tasks in the early perception stage,we use a strategy of sharing weights in the network to fuse cross-domain features,and introduce a novel multi-task loss function to further smooth the depth *** experiments on KITTI and Cityscapes datasets show that our method has achieved state-of-the-art performance in the depth estimation task,as well improved semantic segmentation.

关键词： unsupervised monocular depth estimation semantic segmentation multi-task model

来源：评论

学校读者我要写书评

暂无评论

A Prediction Method for Dimensional Sentiment Analysis of the Movie and TV Drama based on Variable-length Sequence Input

A Prediction Method for Dimensional Sentiment Analysis of th...

引用

2022 International Conference on culture-Oriented Science and technology, CoST 2022

作者： Wang, Chunxiao Zhang, Jingiing Gan, Lihong Jiang, Wei Key Laboratory of Acoustic Visual Technology and Intelligent Control System Communication University of China Beijing China Zhejiang Dafeng sports equipment Co.Ltd Zhejiang China

ISBN: (数字)9781665462488

ISBN: (纸本)9781665462488

Time continuous emotion prediction problem has always been one of the difficulties in affective video content analysis. The current research mainly designs a temporally continuous long video emotion prediction method by dividing the long video into short video segments of fixed duration. These methods ignore the time dependencies between short video clips and the mood changes in short video clips. Therefore, combined with the related concepts of film and television narrative structure in cinematic language, this paper defines a prediction method for dimensional sentiment analysis of the movie and TV drama based on variable sequence length inputs. First, this paper defines a method for partitioning variable-length audiovisual sequences that set subunits of dimensional emotion prediction as variable sequence-length inputs. Then, a method for extracting and combining audio and visual features of each variable-length audiovisual sequence is proposed. Finally, a prediction network for dimensional emotion is designed based on variable sequence length inputs. This paper focuses on dimensional sentiment prediction and evaluates the proposed method on the extended COGNIMUSE dataset. The method achieves comparable performance to other methods while increasing the prediction speed, with the Mean Square Error (MSE) reduced from 0.13 to 0.11 for arousal and from 0.19 to 0.13 for valence. © 2022 IEEE.

关键词： Sentiment analysis

来源：评论

学校读者我要写书评

暂无评论

Simulation of stage dry ice effect based on UE4 6

Simulation of stage dry ice effect based on UE4

引用

6th International Symposium on Computer and Information Processing technology, ISCIPT 2021

作者： Li, Yiyang Li, Jiawen Lv, Chaohui University of China School of Information and Communication Engineering Communication Beijing China Key Laboratory of Acoustic Visual Technology and Intelligent Control System Ministry of Culture and Tourism Beijing China

ISBN: (纸本)9781665441377

In view of the problem that the dry ice effect simulation on the stage simulation is less, a dry ice effect simulation system including the realistic rendering module of dry ice effect and the interaction module between actors and the dry ice is proposed. For visual effect of dry ice, this paper uses a method to simulate volume fog in real time by adding visual optimization of multiple noise mixtures. Firstly, in order to solve the problem that the modeling effect of single noise fog is not natural, the Perlin noise and Worley noise are used to model the fog together to enhance the detail and flip effect of fog. Secondly, using the offset shadow algorithm to simulate the self-shadow effect, combining pixel jitter and Temporal oversampling, the soft effect of fog is increased, and the distance field shadow is added to reflect the occlusion. For dry ice interaction module, this paper uses symbol distance field to detect the collision between dry ice and actors to realize the dynamic interaction between actors and dry ice. The results show that this method can simulate the dry ice effect of stage and has the function of real-time interaction. © 2021 IEEE.

关键词： Ice

来源：评论

学校读者我要写书评

暂无评论

Research on emotional space for movie and TV drama videos

引用

The Journal of China Universities of Posts and Telecommunications 2022年第5期29卷 73-82页

作者： Li Yujie Zhang Jingjing Jiang Wei Wang Chunxiao Key Laboratory of Acoustic Visual Technology and Intelligent Control System Communication University of ChinaBeijing 100024China Beijing Key Laboratory of Modern Entertainment Technology Communication University of ChinaBeijing 100024China School of Information and Communication Engineering Communication University of ChinaBeijing 100024China School of Computer and Cyber Sciences Communication University of ChinaBeijing 100024China

Emotional space refers to a multi-dimensional emotional model that describes a group of subjective feelings or emotions. Since the existing discrete emotional space is mainly aimed at human’s primary emotions, it cannot describe the complex emotions evoked when watching movies. In order to solve this problem, an emotional fusion space for videos was constructed by selecting movies and TV dramas with rich emotional semantics as the research objects. Firstly, emotional words based on movie and TV drama videos are acquired and analyzed by using subjective evaluation and semantic analysis methods. Then, the emotional word vectors obtained from the above analysis are fused, reduced dimension by t-distributed stochastic neighbor embedding(t-SNE) algorithm, and clustered by bisecting K-means clustering algorithm to get a discrete emotional space for movie and TV drama videos. This emotional fusion space can obtain different categories by changing the value of the emotion classification number without re-labeling and calculation.

关键词： emotional space movie and TV drama videos subjective evaluation words semantic analysis fusion space

来源：评论

学校读者我要写书评

暂无评论

Association Between Timbre Perception Features and Fine-Grained Emotions in Film and Television Music

Association Between Timbre Perception Features and Fine-Grai...

引用

culture-Oriented Science and technology (CoST), International Conference on

作者： Xiaomeng Ren Zhibin Su Wei Jiang Jingyu Liu State Key Laboratory of Media Convergence and Communication (Communication University of China) China Key Laboratory of Acoustic Visual Technology and Intelligent Control System Ministry of Culture and Tourism (Communication University of China) China School of Information and Communication Engineering (Communication University of China) China

ISBN: (数字)9798350380347

ISBN: (纸本)9798350380354

As an integral component of modern cultural works, the role of film and television music in artistic expression and emotional communication has become increasingly significant. Timbre, as a crucial attribute of music, has primarily been studied in terms of the influence of its fundamental objective parameters on coarse-grained emotions. However, there is a lack of in-depth exploration of the association between timbre perception features and fine-grained emotion. Consequently, this paper approaches from the perspective of timbre perception features and fine-grained emotions in film and television music. By conducting perceptual evaluation and subjective emotional annotation experiments, it analyzes the correlation between timbre perception features and fine-grained emotions, confirming that timbre perception features significantly influence musical emotions. Furthermore, a multivariate linear regression model was employed to construct a fine-grained emotion prediction model, which has demonstrated good predictive ability for most musical emotions. This research provides a new perspective on understanding the relationship between timbre and emotion in film and television music, offering a scientific basis for its creation and performance.

关键词： TV Costs Correlation Annotations Linear regression Predictive models Timbre

来源：评论

学校读者我要写书评

暂无评论

Dunhuang murals image inpainting method based on fusion classification pre-training model

Dunhuang murals image inpainting method based on fusion clas...

引用

2024 International Conference on culture-Oriented Science and technology, CoST 2024

作者： Ren, Hui Zhang, Chengya Sun, Ke Ye, Chenyu Communication University of China School of Information and Communication Engineering Beijing100024 China Ministry of Culture and Tourism Key Laboratory of Audiovisual Technology and Intelligent Control System Beijing100024 China Beijing Key Laboratory of Modern Performing Arts Technology Beijing100024 China University of Pennsylvania School of Engineering and Applied Science PA19104 United States

ISBN: (纸本)9798350380347

Murals are an important part of China's cultural heritage. Because of their age, Dunhuang murals have suffered from discoloration, fading and damage. With the development of computer image restoration technology, digital image processing provides strong support for mural restoration. Although there are many research results in the field of ancient mural images, there is no pre-trained model based on ancient mural images available for migration. There is a method for Dunhuang murals image restoration based on Generative Adversarial Network (GAN) with improved ResNet50 structure as an encoder in the generator. This research result proposes a generative adversarial network model with a parallel double convolutional extraction of deep features for the generator, combined with a ternary heterogeneous joint discriminator. On this basis, this paper proposes a binary classification task to distinguish ancient mural images from modern natural style images, so that the model pre-learns features such as color, texture, and structure of mural images, which are inputted into the restoration network as a priori information, in order to achieve the effect of reducing the probability of incomplete restoration of the model, and better generalization ability. The experimental results show that the method proposed in this paper has an overall improvement in the evaluation indexes such as Peak Signal to Noise Ratio(PSNR) and Structural Similarities(SSIM) when compared with other methods, and the restored images are more in line with human subjective vision, which proves the validity of the method. © 2024 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Multipath planning based on multimodal optimization slime mold algorithm

Multipath planning based on multimodal optimization slime mo...

引用

culture-Oriented Science and technology (CoST), International Conference on

作者： Luli Gao Hui Ren School of Information and Communication Engineering Communication University of China Bejing China Key laboratory of Acoustic Visual Technology and Intelligent Control System(Communication University of China)Ministry of Culture and Tourism Bejing China Beijing Key Laboratory of Modern Entertainment Technology (Communication University of China) Bejing China

ISBN: (数字)9798350380347

ISBN: (纸本)9798350380354

Multipath planning aims to find multiple optimal paths simultaneously, and its essence is a multimodal optimization problem. In recent years, swarm intelligence (SI) optimization algorithms have provided effective solutions for multimodal optimization. In this paper, the multimodal optimization slime mold algorithm (MOSMA) based on master-slave structure and local search strategy (LS) is proposed to solve multipath planning problems in small cultural complex. Firstly, a local search strategy is introduced to eliminate the influence of the “centerbias” operator and improve the quality of solutions. Secondly, the master-salves structure is introduced to preserve the diversity of the population and enable it to simultaneously locate multiple solutions. Finally, the performance of MOSMA is analyzed on several multimodal functions, and an example of path planning in a small cultural complex is given.

关键词： Location awareness Costs Search problems Main-secondary Path planning Planning Cultural differences

来源：评论

学校读者我要写书评

暂无评论

Obstacle Avoidance Algorithm of Handling Robot in Cultural Complex

Obstacle Avoidance Algorithm of Handling Robot in Cultural C...

引用

2022 International Conference on culture-Oriented Science and technology, CoST 2022

作者： Ren, Hui Li, Mengnan Shen, Xiaochen School of Information and Commmunication Engineering Communication University of China Beijing100024 China Key Laboratory of Acoustic Visual Technology and Intelligent Control System Beijing100024 China Beijing100024 China

ISBN: (数字)9781665462488

ISBN: (纸本)9781665462488

In order to solve the problem of rapid switching of space equipment in the cultural complex, this paper combines Moth-Flame optimization Algorithm (MFO) and Dynamic Window Approach (DWA) to realize the hybrid path planning of handling robots in dynamic environments. Compared with the mainstream algorithms, the Moth-Flame optimization algorithm has the best performance. Different paths were classified and judged, and the global path planning based on MFO was proposed. The intersection of the dynamic obstacle and the global path were detected, and the dynamic window approach was used for local path planning in the road section where the collision occurs. Finally the combination of the global path planning and the local dynamic obstacle avoidance was realized. The effect of the hybrid path planning algorithm in the complex environment of the cultural complex had been verified. © 2022 IEEE.

关键词： Optimization

来源：评论

学校读者我要写书评

暂无评论

Automatic Image Generation of Peking Opera Face using StyleGAN2

Automatic Image Generation of Peking Opera Face using StyleG...

引用

2022 International Conference on culture-Oriented Science and technology, CoST 2022

作者： Xin, Xiaoyu Shen, Yinghua Xiong, Rui Lin, Xiahan Yan, Ming Jiang, Wei Communication University of China School of Information and Communication Engineering Beijing China Center for Ethnic and Folk Literature and Art Development Ministry of Culture and Tourism of the P.R.C Beijing China Communication University of China Key Laboratory of Acoustic Visual Technology and Intelligent Control System Beijing China

ISBN: (数字)9781665462488

ISBN: (纸本)9781665462488

Image generation technology, which is often used in various applications of intelligent image generation, can learn the feature distribution of real images and sample from the distribution to obtain the generated images with high fidelity. This paper focuses on the feature extraction and intelligent generation techniques of Peking opera face with Chinese cultural characteristics. Based on the creation of a Peking opera face dataset, this paper compares the impact of different variants of a Style-based generator architecture for Generative Adversarial Networks (StyleGAN2) and different sizes of datasets on the quality of face generation. The experimental results verify that the synthetic images generated by StyleGAN2 with the addition of the Adaptive Discriminator Augmentation (ADA) module are visually better and have good local randomness when the dataset is small and unbalanced in distribution. © 2022 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：