NAAI International Artificial Intelligence Seminar (Phase 7) – Call for Participation

NAAI International Artificial Intelligence Seminar (Phase 7) – Call for Participation

The National Academy of Artificial Intelligence (NAAI) will hold the 7th International AI Seminar on August 15, 2025, at 9:00 AM (New York Time), featuring distinguished experts sharing insights on Computer Vision and General Artificial Intelligence and GAN-based Anomaly Detection Using Machine Sounds.

Date & Time:

August 15, 2025 (Friday)

9:00 AM, New York Time

Format:

Online via Zoom

Zoom Link: https://us06web.zoom.us/j/81124095140? pwd=R7Kh2A8bt5OEed7ekRb671X2kl0Dc7.1

Meeting ID: 811 2409 5140

Passcode: 982330

Keynote Speakers:

  1. Prof. Sergey Ablameyko

    Member, US-NAAI (National Academy of Artificial Intelligence, USA)

    IAPR Fellow

    Member of Academia Europaea

    Academician of National Academy of Belarus

    Academician of Belarusian Engineering Academy

    Minsk, Belarus
    Topic: Computer Vision and General Artificial Intelligence: Relations, Results and Prospects
    Abstract: Artificial Intelligence aims to build machines capable of performing tasks requiring human-level intelligence. Computer vision, as a crucial AI subfield, has seen significant progress with convolutional neural networks (CNNs). This talk compares classical and deep learning-based image analysis, showcases applications in remote sensing image analysis and video tracking, and discusses the opportunities and challenges of Large Multimodal Models (LMMs) toward General Artificial Intelligence.

  2. Prof. Pingyi Fan

    Member, US-NAAI (National Academy of Artificial Intelligence, USA)

    IET Fellow

    Department of Electronic Engineering, Tsinghua University, Beijing
    Topic: Identifying Machines with Sounds: Anomaly Detection with GAN-based Technology
    Abstract: In the context of Digital Twins and Industry 4.0, anomaly detection is critical for manufacturing and production management. Unlike image-based inspections, audio signals can reveal a machine’s internal conditions. This talk presents AEGAN, a hybrid of GAN and autoencoder, which detects anomalies via generator reconstruction errors and discriminator embedding features, achieving state-of-the-art unsupervised performance on multiple DCASE datasets. The MIM-GAN technique for rare event detection will also be introduced.

Organizer: National Academy of Artificial Intelligence (NAAI)