Proceedings of the First International Conference on Science, Engineering and Technology Practices for Sustainable Development, ICSETPSD 2023, 17th-18th November 2023, Coimbatore, Tamilnadu, India

Research Article

Film and Television Production AI Intelligent Optimization Processing System Based on Neural Optimization Algorithm

Download91 downloads
  • @INPROCEEDINGS{10.4108/eai.17-11-2023.2342814,
        author={Dongsheng  Yang},
        title={Film and Television Production AI Intelligent Optimization Processing System Based on Neural Optimization Algorithm},
        proceedings={Proceedings of the First International Conference on Science, Engineering and Technology Practices for Sustainable Development, ICSETPSD 2023, 17th-18th November 2023, Coimbatore, Tamilnadu, India},
        publisher={EAI},
        proceedings_a={ICSETPSD},
        year={2024},
        month={1},
        keywords={neural optimization algorithm ai film and television production},
        doi={10.4108/eai.17-11-2023.2342814}
    }
    
  • Dongsheng Yang
    Year: 2024
    Film and Television Production AI Intelligent Optimization Processing System Based on Neural Optimization Algorithm
    ICSETPSD
    EAI
    DOI: 10.4108/eai.17-11-2023.2342814
Dongsheng Yang1,*
  • 1: School of Communication, Qufu Normal University, Rizhao 276800, Shandong, China
*Contact email: 17861013177@163.com

Abstract

In the context of the development of ultra high-definition (4K, 8K) industry, artificial intelligence technology represented by deep learning is developing rapidly in the field of image super resolution. Based on the adversarial generation super resolution network (SRGAN), we propose a novel image super resolution generation model combining semantic segmentation probability graph and iterative check kernel (IKC) technology. This model can recognize the target object in the image according to the application requirements and make the generated ultra HD image texture more real. Therefore, we made Images From TV (IFTV) data set based on the big data of radio and television media assets to optimize and train the common application scenes of radio and television (such as those with more faces or text), so that the model could achieve satisfactory super resolution effect in multiple scenes. It will provide strong support for the production of ultra HD content in the field of broadcasting and television in the future.