Report

Mahinur

April 15, 2024 - April 19, 2024

Papers

Paper 1: Advanced Education and Evaluation using Token-based E-Learning Model Harnessing Hyperledger Besu
Target: ICMIC 2024

Paper 2: Anomaly Detection in management execution systems (intuitive)

Paper 3: Collaborative Learning based Real and AI-generated image detection (intuitive)
Target: KICS Summar

40% Progress
Last Week's Progress

- Writing Paper 1 methodology part with results, system model, and deciding title.
- Implementation of the model for anomaly detection (Autoencoder & Transfer Learning such as ResNet50, DenseNet121, Xception, Inceptionv3, EfficientNet).
- Learning about Federated Learning, Secure Collaborative Learning, anomaly detection, image detection, and AI-Image generation
- Reading papers on anomaly detection from images, token-based learning, real and AI-generated image detection, secure collaborative learning, federated learning

This Week Todo's

- Completing the methodology part and intro, related work, and conclusion references.
- Applying anomaly detection models on the Mvtech Loco Ad dataset
- Applying the image classification model to the CIFAKE dataset to detect AI-generated image
- Reading papers on anomaly detection from images, token-based learning, real and AI-generated image detection, secure collaborative learning, federated learning

Project Progress

Project 1: Manufacturing Execution Systems

Project 2: Data Distribution Services (DDSs)

90% Progress
Last Week's Project

- Trying new models (Autoencoder & Transfer Learning such as ResNet50, DenseNet121, Xception, Inceptionv3, EfficientNet)) and data augmentation on the previous model.
- Implementing a simple DDS node to demonstrate data transmission

This Week's Project

- Working on the Solar PV dataset to detect anomalies using VGG16
- Studying the licensing and usability of open-source models like VGG16, and YOLOv8
- Trying new models (Autoencoder & Transfer Learning such as ResNet50, DenseNet121, Xception, Inceptionv3, EfficientNet)) and data augmentation on the previous model

Monthly Goals

- Learning Deep Learning, Autoencoder, Hyperledger Fabric & Besu, DDS.

Annual Goals

- Publish at least one Q1 International Journal.
- Attend international conferences and domestic conferences.