检索结果-南通市图书馆

Dual attention autoencoder for all-weather outdoor lighting estimation

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2021年第11期64卷 22-35页

作者： Piaopiao YU Jie GUO Longhai WU Cheng ZHOU Mengtian LI Chenchen WANG Yanwen GUO National Key Lab for Novel Software technology Nanjing University Samsung Electronics (China)R&D Center kuaishou technology

Accurate outdoor illumination estimation is not easy due to extremely complicated sky appearance and the mutual interference of the sun and sky. In view of these challenges, we present a new deep approach for the estimation of all-weather outdoor illumination. The key to our approach is a novel dual attention autoencoder(DAA) with two independent branches to compress the sun and sky lighting information from an input HDR panorama, respectively. This enables more accurate lighting estimation as evidenced by our experiments since the mutual interference between the sun and sky can be precluded effectively. In DAA,we design the adaptive feature pyramid and the attention module to promote its accuracy in *** further develop a sun-sky predictor, a masked network, to learn the sun and sky lighting conditions from an Fo V-limited image. Comprehensive qualitative and quantitative experiments verify the effectiveness of our proposed approach and show its superiority over the state of the arts.

关键词： outdoor illumination adaptive feature pyramid attention autoencoder augmented reality

HDR-Net-Fusion:Real-time 3D dynamic scene reconstruction with a hierarchical deep reinforcement network

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Computational Visual Media 2021年第4期7卷 419-435页

作者： Hao-Xuan Song Jiahui Huang Yan-Pei Cao Tai-Jiang Mu BNRist Department of Computer Science and TechnologyTsinghua UniversityBeiing 100084China kuaishou technology Co. Ltd.Beijing 100085China

Reconstructing dynamic scenes with commodity depth cameras has many applications in computer graphics,computer vision,and ***,due to the presence of noise and erroneous observations from data capturing devices and the inherently ill-posed nature of non-rigid registration with insufficient information,traditional approaches often produce low-quality geometry with holes,bumps,and *** propose a novel 3D dynamic reconstruction system,named HDR-Net-Fusion,which learns to simultaneously reconstruct and refine the geometry on the fly with a sparse embedded deformation graph of surfels,using a hierarchical deep reinforcement(HDR)*** latter comprises two parts:a global HDR-Net which rapidly detects local regions with large geometric errors,and a local HDR-Net serving as a local patch refinement operator to promptly complete and enhance such *** the global HDR-Net is formulated as a novel reinforcement learning problem to implicitly learn the region selection strategy with the goal of improving the overall reconstruction *** applicability and efficiency of our approach are demonstrated using a large-scale dynamic reconstruction *** method can reconstruct geometry with higher quality than traditional methods.

关键词： dynamic 3D scene reconstruction deep reinforcement learning point cloud completion deep neural networks

维普期刊数据库

EM-Gaze:eye context correlation and metric learning for gaze estimation

在线全文

学校读者我要写书评

暂无评论

Visual Computing for Industry,Biomedicine,and Art 2023年第1期6卷 97-108页

作者： Jinchao Zhou Guoan Li Feng Shi Xiaoyan Guo Pengfei Wan Miao Wang State Key Laboratory of Virtual Reality technology and Systems Beihang UniversityBeijing 100191China kuaishou technology Beijing 100085China

In recent years,deep learning techniques have been used to estimate gaze-a significant task in computer vision and human-computer *** studies have made significant achievements in predicting 2D or 3D gazes from monocular face *** study presents a deep neural network for 2D gaze estimation on mobile *** achieves state-of-the-art 2D gaze point regression error,while significantly improving gaze classification error on quadrant divisions of the *** this end,an efficient attention-based module that correlates and fuses the left and right eye contextual features is first proposed to improve gaze point regression ***,through a unified perspective for gaze estimation,metric learning for gaze classification on quadrant divisions is incorporated as additional ***,both gaze point regression and quadrant classification perfor-mances are *** experiments demonstrate that the proposed method outperforms existing gaze-estima-tion methods on the GazeCapture and MPIIFaceGaze datasets.

关键词： Computer vision Gaze estimation Metric learning Attention Multi-task learning

Review of light field technologies

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Visual Computing for Industry,Biomedicine,and Art 2021年第1期4卷 295-307页

作者： Shuyao Zhou Tianqian Zhu Kanle Shi Yazi Li Wen Zheng Junhai Yong Y-tech Kuaishou TechnologyBeijing 100085China EECS Department University of CaliforniaBerkeleyCA 94720USA School of Software BNRistTsinghua UniversityBeijingChina

Light fields are vector functions that map the geometry of light rays to the corresponding plenoptic *** describe the holographic information of scenes by representing the amount of light flowing in every direction through every point in *** physical concept of light fields was first proposed in 1936,and light fields are becoming increasingly important in the field of computer graphics,especially with the fast growth of computing capacity as well as network *** this article,light field imaging is reviewed from the following aspects with an emphasis on the achievements of the past five years:(1)depth estimation,(2)content editing,(3)image quality,(4)scene reconstruction and view synthesis,and(5)industrial products because the technologies of lights fields also intersect with industrial ***-of-the-art research has focused on light field acquisition,manipulation,and *** addition,the research has extended from the laboratory to *** to these achievements and challenges,in the near future,the applications of light fields could offer more portability,accessibility,compatibility,and ability to visualize the world.

关键词： Light field imaging Holographics Human-machine graphic interaction

Emotion-Aware Music Driven Movie Montage

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Computer Science & technology 2023年第3期38卷 540-553页

作者：刘伍琴林敏轩黄海斌马重阳宋玉董未名徐常胜 School of Artificial Intelligence University of Chinese Academy of SciencesBeijing 101408China The State Key Laboratory of Multimodal Artificial Intelligence System(MAIS) Institute of AutomationChinese Academy of SciencesBeijing 100190China kuaishou technology Beijing 100085China School of Mechanical Engineering University of Science and Technology BeijingBeijing 100083China

In this paper, we present Emotion-Aware Music Driven Movie Montage, a novel paradigm for the challenging task of generating movie montages. Specifically, given a movie and a piece of music as the guidance, our method aims to generate a montage out of the movie that is emotionally consistent with the music. Unlike previous work such as video summarization, this task requires not only video content understanding, but also emotion analysis of both the input movie and music. To this end, we propose a two-stage framework, including a learning-based module for the prediction of emotion similarity and an optimization-based module for the selection and composition of candidate movie shots. The core of our method is to align and estimate emotional similarity between music clips and movie shots in a multi-modal latent space via contrastive learning. Subsequently, the montage generation is modeled as a joint optimization of emotion similarity and additional constraints such as scene-level story completeness and shot-level rhythm synchronization. We conduct both qualitative and quantitative evaluations to demonstrate that our method can generate emotionally consistent montages and outperforms alternative baselines.

关键词： movie montage emotion analysis audio-visual modality contrastive learning

Learning to assess visual aesthetics of food images

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computational Visual Media 2021年第1期7卷 139-152页

作者： Kekai Sheng Weiming Dong Haibin Huang Menglei Chai Yong Zhang Chongyang Ma Bao-Gang Hu Youtu Lab TencentShanghai 200233China NLPR Institute of AutomationChinese Academy of SciencesBeijing 100190China kuaishou technology Beijing 100085China Snap Inc. Santa Monica90405USA AI Lab Tencent Inc.Shenzhen 518000China

Distinguishing aesthetically pleasing food photos from others is an important visual analysis task for social media and ranking systems related to ***,aesthetic assessment of food images remains a challenging and relatively unexplored task,largely due to the lack of related food image datasets and practical ***,we present the Gourmet Photography Dataset(GPD),the first largescale dataset for aesthetic assessment of food *** contains 24,000 images with corresponding binary aesthetic labels,covering a large variety of foods and *** also provide a non-stationary regularization method to combat over-fitting and enhance the ability of tuned models to *** results from extensive experiments,including a generalization ability test,verify that neural networks trained on the GPD achieve comparable performance to human experts on the task of aesthetic *** reveal several valuable findings to support further research and applications related to visual aesthetic analysis of food *** encourage further research,we have made the GPD publicly available at https://***/Openning07/GPA.

关键词： image aesthetic assessment food image analysis dataset regularization

维普期刊数据库

A review of feature fusion-based media popularity prediction methods

在线全文

学校读者我要写书评

暂无评论

Visual Informatics 2022年第4期6卷 78-89页

作者： An-An Liu Xiaowen Wang Ning Xu Junbo Guo Guoqing Jin Quan Zhang Yejun Tang Shenyuan Zhang State Key Laboratory of Communication Content Cognition People’s Daily OnlineBeijing100733China School of Electrical and Information Engineering Tianjin UniversityTianjin300072China Peking University Beijing100871China kuaishou technology Beijing100000China New Media Center People’s DailyBeijing100000China

With the popularization of social media,the way of information transmission has changed,and the prediction of information popularity based on social media platforms has attracted extensive *** fusion-based media popularity prediction methods focus on the multi-modal features of social media,which aim at exploring the key factors affecting media ***,the methods make up for the deficiency in feature utilization of traditional methods based on information propagation *** this paper,we review feature fusion-based media popularity prediction methods from the perspective of feature extraction and predictive model *** that,we analyze the influencing factors of media popularity to provide intuitive *** further argue about the advantages and disadvantages of existing methods and datasets to highlight the future ***,we discuss the applications of popularity *** the best of our knowledge,this is the first survey reporting feature fusion-based media popularity prediction methods.

关键词： Social media Popularity prediction Multi-modal analysis

Study on the application of reinforcement learning in the operation optimization of HVAC system

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Building Simulation 2021年第1期14卷 75-87页

作者： Xiaolei Yuan Yiqun Pan Jianrong Yang Weitong Wang Zhizhong Huang School of Mechanical Engineering Tongji University4800 Cao’an RoadShanghai201804China Shanghai Research Institute of Building Sciences ShanghaiChina kuaishou Co.Ltd. BeijingChina Sino-German College of Applied Sciences Tongji UniversityShanghai201804China

Supervisory control can be used to optimize the HVAC system operation and achieve building energy conservation,while reinforcement learning(RL)is considered as a promising model-free supervisory control *** this paper,we apply RL algorithm to the operation optimization of air-conditioning(AC)system and propose an innovative RL-based model-free control strategy combining rule-based and RL-based control algorithm as well as complete application *** use a variable air volume(VAV)air-conditioning system for a single-storey office building as a case study to validate the optimization performance of the RL-based *** select control strategies with the rule-based control controller(RBC)and proportional-integral-derivative(PID)controller respectively as the reference *** results show that,for the air supply of single zone,the RL controller performs the best in terms of both non-comfortable time and energy costs of AC system after one-year exploration *** total energy consumption of AC system reduced by 7.7%and 4.7%,respectively compared with RBC and PID *** the air supply of multi-zone,the performance of RL controller begins to outperform the reference strategies after two-year exploration learning and two-year buffer *** the seventh year on,RL controller performs much better in terms of both non-comfortable time and operating costs of AC system,while the operating cost of AC system is reduced by 2.7%to 4.6%compared with the reference *** addition,RL controller is more suitable for small-scale operation optimization problems.

关键词： reinforcement learning HVAC system operation optimization control strategy VAV system energy saving

A novel robotic visual perception framework for underwater operation

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Information technology & Electronic Engineering 2022年第11期23卷 1602-1619页

作者： Yue LU Xingyu CHEN Zhengxing WU Junzhi YU Li WEN State Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of SciencesBeijing 100190China Ytech Kuaishou TechnologyBeijing 100085China State Key Laboratory for Turbulence and Complex Systems Department of Advanced Manufacturing and RoboticsCollege of EngineeringPeking UniversityBeijing 100871China School of Mechanical Engineering and Automation Beihang UniversityBeijing 100191China

Underwater robotic operation usually requires visual perception(e.g.,object detection and tracking),but underwater scenes have poor visual quality and represent a special domain which can affect the accuracy of visual *** addition,detection continuity and stability are important for robotic perception,but the commonly used static accuracy based evaluation(i.e.,average precision)is insufficient to reflect detector performance across *** response to these two problems,we present a design for a novel robotic visual perception ***,we generally investigate the relationship between a quality-diverse data domain and visual restoration in detection *** a result,although domain quality has an ignorable effect on within-domain detection accuracy,visual restoration is beneficial to detection in real sea scenarios by reducing the domain ***,non-reference assessments are proposed for detection continuity and stability based on object ***,online tracklet refinement is developed to improve the temporal performance of ***,combined with visual restoration,an accurate and stable underwater robotic visual perception framework is ***-overlap suppression is proposed to extend video object detection(VID)methods to a single-object tracking task,leading to the flexibility to switch between detection and *** experiments were conducted on the ImageNet VID dataset and real-world robotic tasks to verify the correctness of our analysis and the superiority of our proposed *** codes are available at https://***/yrqs/VisPerception.

关键词： Underwater operation Robotic perception Visual restoration Video object detection

维普期刊数据库