Showing papers on "Multimedia database published in 2018"

DeepStyle: Multimodal Search Engine for Fashion and Interior Design

Abstract: In this paper, we propose a multimodal search engine that combines visual and textual cues to retrieve items from a multimedia database aesthetically similar to the query. The goal of our engine is to enable intuitive retrieval of fashion merchandise such as clothes or furniture. Existing search engines treat textual input only as an additional source of information about the query image and do not correspond to the real-life scenario where the user looks for 'the same shirt but of denim'. Our novel method, dubbed DeepStyle, mitigates those shortcomings by using a joint neural network architecture to model contextual dependencies between features of different modalities. We prove the robustness of this approach on two different challenging datasets of fashion items and furniture where our DeepStyle engine outperforms baseline methods by 18-21% on the tested datasets. Our search engine is commercially deployed and available through a Web-based application.

A modified K-means clustering for mining of multimedia databases based on dimensionality reduction and similarity measures

Abstract: With rapid innovations in digital technology and cloud computing off late, there has been a huge volume of research in the area of web based storage, cloud management and mining of data from the cloud Large volumes of data sets are being stored, processed in either virtual or physical storage and processing equipments on a daily basis Hence, there is a continuous need for research in these areas to minimize the computational complexity and subsequently reduce the time and cost factors The proposed research paper focuses towards handling and mining of multimedia data in a data base which is a mixed composition of data in the form of graphic arts and pictures, hyper text, text data, video or audio Since large amounts of storage are required for audio and video data in general, the management and mining of such data from the multimedia data base needs special attention Experimental observations using well known data sets of varying features and dimensions indicate that the proposed cluster based mining technique achieves promising results in comparison with the other well-known methods Every attribute denoting the efficiency of the mining process have been compared component wise with recent mining techniques in the past The proposed system addresses effectiveness, robustness and efficiency for a high-dimensional multimedia database

Multimodal query‐level fusion for efficient multimedia information retrieval

Abstract: Managing a large volume of multimedia data containing various modalities such as visual, audio, and text reveals the necessity for efficient methods for modeling, processing, storing, and retrieving complex data. In this paper, we propose a fusion‐based approach at the query level to improve query retrieval performance of multimedia data. We discuss various flexible query types including the combination of content as well as concept‐based queries that provide users with the ability to efficiently perform multimodal querying. We have carried out a number of experiments on a video database to show the efficiency of our approach for various types of queries. Our experimental results show that our query‐level fusion approach presents a notable improvement in retrieval performance especially for the concept‐based queries.

Optimal k-Nearest-Neighbor Query Processing via Multiple Lower Bound Approximations

Abstract: Given a very large multimedia database, how to process k-nearest-neighbor queries efficiently? While the sequential scan is one of the most obvious solutions for small-to-moderate multimedia databases, it becomes practically infeasible when the database size grows. Concomitant with the volume and velocity of data, multimedia databases are frequently endowed with a complex distance-based similarity model that supports content-based data access in an adjustable and adaptive manner. Typical for many state-of-the-art distance-based similarity models is an at least quadratic computation time complexity for a single distance evaluation between two multimedia objects. Thus the search for the most query-like multimedia objects is still one of the major challenges.In this paper, we address the problem of optimal k-nearest-neighbor query processing via multiple lower bound approximations in very large multimedia databases. To this end, we propose the concepts of filter-optimality and refinement-optimality and present the Cascading Multi-Step Algorithm and the Interleaved Multi-Step Algorithm for fast query processing. Besides the algorithms’ properties, we study their query processing performance with respect to the number of CPU and I/O operations on large-scale benchmark multimedia databases. Our performance analysis shows how to process k-nearest-neighbor queries in multimedia databases efficiently and provides a guide for further research.

An Image-Based Multimedia Database and Efficient Detection though Features

Abstract: Accurate feature detection during Image retrieval is important, data retrieves through image retrieval methods like CBIR and CBIR higher dimension data also need storage and access through different methods, content-based Image retrieval uses query like query by feature and query by example. More focus has made on accurate feature detection because need accurate feature retrieval. In simple words objectives are, to develop methods with sequence to classify features with normalization for efficient image retrieval from bulk dataset and also to improve method for local and global feature retrieval with automatic feature detection along accuracy. After study of different detection-based system, a methodology has been proposed which improves retrieval based on feature detection and feature detection had been improve with combination DWT+PCA+KSVM (polygon kernel +RBF kernel + Linear Kernel).

Cross-media feature learning retrieval method based on semi-supervision

Abstract: The invention provides a cross-media feature learning retrieval method based on semi-supervision. The method comprises the following steps that: S1: establishing a multimedia database; S2: solving theprojection matrixes of different media types; and S3: carrying out cross-media retrieval; and S3: carrying out cross-media retrieval. The S2 comprises the following steps that: 2.1: defining a targetfunction; 2.2: optimizing the target function; and 2.3: projecting the original feature of the multimedia data to a public space. The S3 comprises the following steps that: 3.1: extracting the feature of the media data submitted by a user: according to the media type of the data submitted by the user, using a pre-trained model to extract the feature of the data; 3.2: projecting the feature vectorof the media data into a common space; 3.3: calculating a similarity between the projected feature vector and other vectors in the common space; and 3.4: returning the first k pieces of media data with the highest similarity. By use of the method, calculation complexity is lowered, noise robustness is realized, and retrieval accuracy is improved.

Multimedia file access control method, terminal and computer readable storage medium

Abstract: The present invention discloses a multimedia file access control method, a terminal, and a computer readable storage medium. The method comprises: obtaining an access request of an application to a multimedia database; parsing the access request and obtaining the type information of each target multimedia file to be accessed by the application according to the parsing result; according to type information of each target multimedia file and the correspondence between the type information of a preset multimedia file and access permission information, inquiring the access permission information corresponding to each target multimedia file; inquiring record information of each target multimedia file with the first permission information in the multimedia database; and feeding back the record information of the target multimedia file to the application. According to the technical scheme of the present invention, for multimedia files with different types, the access permission is controlledrespectively, and user experience is effectively enhanced.

Multimedia Data Modeling and Management

Abstract: Nowadays, multimedia data is surely one of the most popular and pervasive information and communication media that accompanies us in almost every walk of lives. They allow fast and effective communication and sharing of information about peoples’ lives, their behaviors, works, interests, and they are also the digital testimony of facts, objects, and locations and have become an essential component of social media networks. Technically speaking, how to organize and structure this huge amount of data using different paradigms, so that we can easily get useful information, has been a challenging research field for decades. In this chapter we will describe the main results produced by the Multimedia Database Research Group of University of Naples in this area: models for representing multimedia data and the related knowledge and techniques for their storage, indexing and retrieval. In addition, we also point out several applications, with a particular emphasis on social media networks.

Photo album generation method and terminal

Abstract: The invention discloses a photo album generation method and terminal. The photo album generation method comprises the steps that target geographic location is obtained; a multimedia database is searched according to the target geographic location to find multimedia data corresponding to the target geographic location, wherein the multimedia data comprise at least one of image data, video data andvoice data; and the multimedia data automatically generate a multimedia photo album. According to the terminal, the target geographic location is obtained, the generation location of the multimedia data in the terminal is extracted, it is judged whether the generation location of the multimedia data and the target geographic location are in the same area or not, multimedia files of which the generation location and target geographic location belong to the same area are taken as the multimedia data corresponding to the target geographic location, the multimedia photo album is generated based on the multimedia data corresponding to the target geographic location, and thereby a user can conveniently view the multimedia data corresponding to the target geographic location.

System and method for customizing a presentation

Abstract: A technique is provided for customizing a presentation. The technique includes recording multimedia corresponding to a presenter of a presentation. The recorded multimedia is analyzed to extract a representative information corresponding to the multimedia. Further, one or more pre-recorded multimedia files are determined from a multimedia database. The determination is based on a comparison of the representative information with one or more tags associated with each of a plurality of pre-recorded multimedia files. Subsequently, the presentation is customized by inserting the one or more pre-recorded multimedia in the presentation.

Cross-media retrieval method based on subspace learning and semi-supervised regularization

Abstract: The invention provides a cross-media retrieval method based on subspace learning and semi-supervised regularization, which is characterized by comprising the following steps: Step 1, establishing a multimedia database, collecting multimedia original data, extracting features of multimedia data, and storing feature vectors of the multimedia data and the original data; Step 2, acquiring a projectionmatrix of different media types, defining an optimal target function, solving the optimal target function by utilizing an iterative method, and projecting the feature vectors of the multimedia data to a public space; Step 3, carrying out cross-media retrieval, extracting features of media data submitted by a user, projecting feature vectors of the media data into the public space, calculating a similarity between the projected vectors and other vectors in the public space, and returning the media data corresponding to first k feature vectors with the largest similarity with other vectors in the public space. The cross-media retrieval method provided by the invention generates a more accurate retrieval result.

Multimedia-data recording method, terminal and computer-readable storage medium

Abstract: The invention discloses a multimedia-data recording method, a terminal and a computer-readable storage medium. The method includes: before scanning of multimedia data is carried out, accessing presetlocations in a memory to query whether recording information of the to-be-scanned multimedia data already exists at the preset locations, wherein access efficiency which the recording information thatis of the multimedia data and stored at the preset locations has is higher than access efficiency which the recording information that is of the multimedia data and stored in a multimedia database has; and if the recording information does not exist at the preset locations, scanning the to-be-scanned multimedia data, and writing generated recording information of the to-be-scanned multimedia datainto the multimedia database. The terminal and the computer-readable storage medium disclosed by the invention adopt a principle which is similar to the above-mentioned method. According to the method, scanning time can be reduced, consumption on terminal performance can be reduced, scanning can be very quickly completed, thus the problem that the multimedia data cannot be invoked due to that scanning is not completed can be effectively avoided, and user experience is improved.

Data mining method based on multimedia database

Abstract: The invention relates to a data mining method based on a multimedia database. The method includes the steps of constructing a system prototype for mining knowledge, adopting a mode of extracting hidden knowledge or other non-explicit storage from unstructured or semi-structured multimedia data, and organically and creatively combining the multimedia data modeling representation, storage and retrieval and other multimedia database technologies with data mining technologies in the field of relational databases. An MDMP is used as the prototype to achieve a multi-level and multi-grade mining technology. The MDMP uses an MDB as a data platform, utilizes content-based retrieval and related data collection based on user requests for creating a cube of media data features and mining implied rules, and explains acquired knowledge to users on a graphical interface.

Noise Models to Detect Noisy Multimedia Data & its De-Noising

Abstract: The Multimedia database is a collection of relative multimedia data, which includes primary media data types like text, image, and graphical objects like drawing, sketches and illustrations, animation sequences, audio and video, etc. Noise in multimedia data is nothing but an unwanted information included in data that decreases the quality of multimedia data. Noise gets included in multimedia data during multimedia acquisition, transmission, coding or processing steps. De-noising process restores the details of original image as much as possible. Noise removal algorithm selected based on type of noise multimedia data. It is difficult to elaborate and perform denoising actions without prior knowledge of the noise models. This paper presents overview of various noise models, the results of application of different noise models, also focuses of applying different filters for image de-noising.

Multimedia document broadcast control system and multimedia system broadcast control method

Abstract: The invention discloses a multimedia document broadcast control system and a multimedia system broadcast control method. The disclosed multimedia document broadcast control system comprises a server and multiple terminals, wherein the server is used for delivering a multimedia document distribution instruction to a terminal, receiving a multimedia document download request, and then transmitting amultimedia document to the terminal, and the terminal is used for receiving the multimedia document distribution instruction, then sending the multimedia document download request to the server, andsaving the multimedia document in a multimedia database, and is further used for playing a played object. According to the disclosed multimedia system broadcast control method, the instruction is delivered so as to enable the terminal to initiate the download request, and the server downloads the multimedia document. Through adoption of the multimedia document broadcast control system and the multimedia system broadcast control method, the multimedia document is distributed efficiently and timely.

Play control method of multimedia data, play control device of multimedia data and terminal

Abstract: The embodiment of the invention discloses a play control method of multimedia data, a play control device of multimedia data and a terminal, wherein the method comprises the following steps: calling asentiment classification model to perform classification analysis for user information, and determining a target sentiment class which a user object belongs to; according to grade feature reference information associated with the target sentiment class, identifying feature indication information used for representing sentiment degree from the user information, and determining a target strength grade of the user object under the target sentiment class according to the feature indication information; acquiring multimedia data associated with the target strength grade from a multimedia database;and finely controlling play of the multimedia data according to the sentiment class of the user and the grade level finely divided under the sentiment class.

Multimedia transmission monitoring management system

Abstract: The invention relates to a multimedia transmission monitoring management system. Multimedia stored in a multimedia provision center is transmitted to an electronic apparatus through a network system,so that the electronic apparatus analyzes the received multimedia, and a user of the electronic apparatus can obtain required multimedia contents without consuming excessive downloading time. The system comprises a summarization database, a multimedia generation module, a multimedia database and a master tape generation module. Animation files are generated according to object parameters comprisedin master tape data; the storage space occupied by the master tape data does not exceed 10Mbytes, so that the animation files needing to occupy the relatively large storage space do not need to be downloaded, and the file downloading speed is increased. The system can be applied to teaching websites or animation provision websites, thereby increasing the internet access speed of internet users.

A Multimedia Database Server for Content-Based Retrieval and Image Mining

Abstract: This article presents a possible extension of a software tool implemented in C++ that manages multimedia data collections from medical domain. An element of originality for this database management system is that it includes a series of algorithms used for extracting visual information from images (texture and color characteristics) along with classical operations needed for databases servers. It is also presented a data mining algorithm adapted to the database system that will be included in a future version.

Multimedia information flow synchronization device and method

Abstract: The invention relates to a multimedia information flow synchronization device and method, wherein the device is connected to a network, and comprises a detection module, a comparison module, a synchronization module, a storage and a multimedia database, which are coupled with each other. The method is realized through the device, and comprises the steps that: the detection module automatically detects whether a synchronized device exists in the network or not; when the detection module detects the synchronized device, the comparison module compares media indexes of local and long-distance multimedia data, and determines the multimedia data, which is different from the media indexes of the local media data, in the media indexes of the long-distance multimedia data as synchronized multimediadata; the synchronization module downloads the synchronized multimedia data and transmits the synchronized multimedia data to the storage; and the storage stores the synchronized multimedia data andoutputs the media indexes of the synchronized multimedia data to the multimedia database. Due to automatic backup of the multimedia data, users can conveniently use all the multimedia data in the network.

Multimedia playing method and system for moving vehicle

Abstract: A multimedia playing method and a multimedia playing system for moving vehicle are provided. The method includes following steps. At least one multimedia content is stored in a multimedia database. A search condition is obtained according to environmental information nearby the moving vehicle. The multimedia database is queried according to the search condition to retrieve at least one target multimedia content matching to the search condition. The target multimedia content is played by a multimedia playing device on the moving vehicle.

Multimedia Database Server Implementing Content based Retrieval

Abstract: The article presents a software tool implemented in C++ that implements a multimedia database server. An element of originality is that along with the classical functions of a server it has a specialized module for content based retrieval. The users can execute both simple text based queries and complex visual queries, based on a query image. The server processes the images and extracts the color and texture characteristics and stores them in a new data type called IMAGE. The image color information is represented by means of color histograms resulting from the transformation of the RGB color space to HSV color space and the quantization to 166 colors. In order to represent the texture it is considered the co-occurrence matrices .To compute the dissimilitude between the images, the histogram intersection has been used for the color and the Euclidian distance for the texture. It is also presented the client-server communication based on SQL language.

Multimedia playing method and system for mobile carrier

Abstract: The invention provides a multimedia playing method and system for a mobile carrier. The method includes the following steps: storing at least one multimedia content in a multimedia database; acquiringretrieval conditions based on surrounding environmental information of the mobile carrier; querying the multimedia database according to the retrieval conditions to find at least one target multimedia content that meets the retrieval conditions; and playing the target multimedia content through a multimedia playing device on the mobile carrier.

Content-Based Audiovisual Archive Retrieval:Multimedia Database

Abstract: In the last few years extensive request for user oriented multimedia information systems has developed.Multimedia database can be defined as a pool of storage and retrieval systems, in which large amount of media objects are created, searched, modifiedand retrieved. Multimedia is the combination of text, image, graphicsand animations, audio and video information. The addition of database application to handle multimedia objects requires organization of multiple media data streams. Apart from text retrieval, the current waves in web searching and multimedia documents retrieval are the exploration for and supply of images, audio, 3D extracts and video. The content-based multimedia information retrieval offers new techniques and methods for probing various multimedia databases over the world.The discussiondeals with a new standard for multimedia search based on content. Keywords— Multimedia database;content based retrivel; text based retrival; free browsing;CAS;