Fully Automatic CNN Design with Inception Blocks

Togzhan Barakbayeva, M. Fatih Demirci

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    1 Citation (Scopus)

    Abstract

    Convolutional Neural Networks (CNNs) are widely used in image classification tasks and have demonstrated promising classification accuracy results. Designing a CNN architecture requires a manual adjustment of parameters through a series of experiments as well as sufficient knowledge both in the problem domain and CNN architecture design. Therefore, it is extremely difficult for users without prior experience to design a promising CNN for their purposes. To solve this issue, various solutions on the automatic construction of CNN architectures are proposed, including but not limited to a genetic algorithm AE-CNN1 for automatically evolving CNN design using ResNet and DenseNet blocks. In this paper, a significant improvement of the afore-mentioned solution by the addition of Inception blocks is introduced. Performance of the algorithm is assessed on the CINIC-10 benchmark dataset without and with the usage of Inception blocks. As it can be observed from the outcome of the experiment, the addition of Inception blocks positively affects the final classification accuracy. The proposed algorithm does not only improve the current solution but also keeps the advantages of automatic CNN without requesting any manual interventions.

    Original languageEnglish
    Title of host publicationThirteenth International Conference on Digital Image Processing, ICDIP 2021
    EditorsXudong Jiang, Hiroshi Fujita
    PublisherSPIE
    ISBN (Electronic)9781510646001
    DOIs
    Publication statusPublished - 2021
    Event13th International Conference on Digital Image Processing, ICDIP 2021 - Singapore, Singapore
    Duration: May 20 2021May 23 2021

    Publication series

    NameProceedings of SPIE - The International Society for Optical Engineering
    Volume11878
    ISSN (Print)0277-786X
    ISSN (Electronic)1996-756X

    Conference

    Conference13th International Conference on Digital Image Processing, ICDIP 2021
    Country/TerritorySingapore
    CitySingapore
    Period5/20/215/23/21

    Keywords

    • Automatic CNN construction
    • Convolutional neural networks
    • Genetic algorithms

    ASJC Scopus subject areas

    • Electronic, Optical and Magnetic Materials
    • Condensed Matter Physics
    • Computer Science Applications
    • Applied Mathematics
    • Electrical and Electronic Engineering

    Fingerprint

    Dive into the research topics of 'Fully Automatic CNN Design with Inception Blocks'. Together they form a unique fingerprint.

    Cite this