Advanced Detection of AI-Generated Images

Project Code :TCMAPY2028

Objective

This project presents a transformer-enhanced and attention-augmented image classification system designed to distinguish AI-generated images from actual ones using the CIFAKE dataset. It introduces a three-model ensemble combining ResNet-50 with SE attention, EfficientNetV2-S with cross-attention, and MobileNetV3-Large with lightweight ECA attention for refined feature extraction. The system is deployed through a Flask-based interface with modules for registration, login, classification, and logout. The approach offers an efficient, lightweight, and robust pipeline for synthetic image detection, demonstrating strong representation learning through attention-driven architectures.

Abstract

This project focuses on the development of a hybrid deep learning model combined with XGBoost for accurate plant disease detection. By integrating models such as Xception, ResNet50V2, DenseNet121 with XGBoost, the system efficiently classifies plant diseases using image data. The system uses a Flask-based web application for interaction, where users can upload images of plants and receive predictions on the disease category. The dataset used contains a variety of plants and their respective diseases, which the model is trained to identify. By using hybrid techniques that combine deep learning and machine learning, the system enhances prediction accuracy and computational efficiency. This paper discusses the design, implementation, and performance evaluation of the proposed system, with a focus on achieving high accuracy across different plant species and diseases.

Keywords: hybrid model, deep learning, XGBoost, plant disease detection, Flask web app, image classification, CNN, machine learning, feature extraction, model integration.

NOTE: Without the concern of our team, please don't submit to the college. This Abstract varies based on student requirements.