In an age characterized by the relentless growth of multimedia data, the efficient management and processing of video content have emerged as pressing challenges. This paper introduces “EzStream,” a dynamic video-to-text conversion system that transcends the conventional boundaries of compression techniques. Rather than merely reducing file sizes, EzStream delves deeper by converting video content into a textual format, thereby encapsulating the essence of the information embedded within. What sets EzStream apart is its adaptability, as it can be retrained to cater to a diverse array of applications, ensuring that the extracted text aligns precisely with specific task requirements. Our extensive experimentation, grounded in the Human Motion Database (HMDB) dataset, has unveiled a pivotal insight: the selection of an effective network architecture profoundly influences encoding efficiency. We present a comprehensive analysis of EzStream's performance, showcasing the tangible impact of architectural choices on video encoding. Our results underscore a series of transformative benefits, including remarkable reductions in file sizes, expedited processing times, and enhancements in video quality. This paper serves as a beacon, illuminating the indispensable role of network architecture in video processing. It highlights how this architectural foundation empowers EzStream to excel in encoding, ultimately reshaping the landscape of multimedia data management. In an era marked by the exponential growth of multimedia content, the innovation embodied in EzStream not only meets but anticipates the escalating demands for resource-efficient multimedia processing, promising more agile and adaptive solutions for information extraction and utilization.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    EzStream: Video Compression and Textual Representation for Multimedia Sensor Networks


    Contributors:


    Publication date :

    2023-11-22


    Size :

    463632 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Spatiotemporal video representation and compression

    Kuleshov, S. / Zaytseva, A. / Aksenov, A. | British Library Online Contents | 2013


    Sensor assisted video compression

    STEPHAN MARTIN / MOHR ULRICH | European Patent Office | 2016

    Free access

    SENSOR ASSISTED VIDEO COMPRESSION

    STEPHAN MARTIN / MOHR ULRICH | European Patent Office | 2016

    Free access

    Tactical sensor data compression for military multimedia systems

    Williams, Jim / Yang, Chao-Kung | AIAA | 1993


    High-bitrate multimedia information hiding for digital image/video under lossy compression

    Yang, M. / Bourbakis, N. | British Library Online Contents | 2007