Special issue on Hierarchical Representation Learning for Big Complex Multimedia Data
摘要截稿:
全文截稿: 2018-05-31
影响因子: 2.313
期刊难度:
CCF分类: C类
中科院JCR分区:
• 大类 : 计算机科学 - 4区
• 小类 : 计算机:信息系统 - 4区
• 小类 : 计算机:软件工程 - 4区
• 小类 : 计算机:理论方法 - 4区
• 小类 : 工程:电子与电气 - 4区
Overview
Big multimedia data and deep learning are two high-focus of multimedia data study. The sheer volume of multimedia data is growing exponentially due to the availability of ubiquitous and cheap sensors, which also enables both public and private to collect massive amounts of domain-specific information containing useful information about problems such as surveillance intelligence, cyber security, text/image/video understanding in social networks. With the ever-increasing speed of generating, processing, and sharing multimedia data, the necessity of extracting compact representations from large-volume multimedia data is on demand, and will impact existing and future technologies. However, providing solutions to multimedia data in complexity such as unconstrained videos and images captured in the wild and multimodal data brings about a higher level of difficulty at attempting to understand their contents. Also, we need to address some important problems in big multimedia data analytics, including extracting complex patterns from massive volumes of data, semantic indexing/retrieval, fast information retrieval, and simplifying discriminative tasks. Deep learning algorithms extract high-level, complex abstractions as data representations through a hierarchical learning process. Complex abstractions are learned at a given level based on relatively simpler abstractions formulated in the preceding level in the hierarchy. A key benefit of deep learning is the analysis and learning of massive amounts of unsupervised data, making it valuable for big complex multimedia analytics where raw multimedia data is often yet largely unlabeled and non-categorized.
This special issue is committed to focus on the most recent progress in deep learning techniques in solving complex multimedia data modeling and understanding. We aim at encouraging original research and promoting activities contributive to different types of cutting-edge techniques towards big complex multimedia data systems. The primary objective of this special issue is to attract focused attention on the latest research progress in this emerging area. The list of potential topics includes but not limited to:
· Supervised/unsupervised deep feature learning
· Multimedia data tagging
· Data security issues for multimedia data including forensic watermarking, forensic utilization of biometrics
· High-dimensional data embedding and modeling
· Content analysis and mining for multimedia big data
· Semantic indexing and hashing
· Semantic retrieval of multimedia data
· Distributed and parallel computing in big multimedia analytics
· Data preparation and representation extraction for discriminative and indexing tasks
· Applications of deep learning in big multimedia data analytics
· Trust and privacy issues in big multimedia data systems
· Learning invariant feature representations from multi-modal data
· Deep learning and cloud computing for complex multimedia data
· Data collections, benchmarking and performance evaluations
· Security and privacy applications in multimedia systems
· Innovative and incentive applications in big complex multimedia data in various fields (e.g., search, health care, marketing, medical informatics)