UCI-SAP-Capstone-2024 / Proof-of-Performance-Fraud-Detection

0 stars 1 forks source link

Verifying accuracy of ML models on AI generated images for custom dataset #24

Open NeilNagaraj opened 7 months ago

NeilNagaraj commented 7 months ago

Working on ML model called 'AI vs Real image detection on Hugging face' (https://huggingface.co/dima806/ai_vs_real_image_detection). This uses CNN and uses a dataset from kaggle called 'Cifake AI generated Image detection'(https://www.kaggle.com/code/dima806/cifake-ai-generated-image-detection-vit). This model as of now gives an accuracy of about 80% but the accuracy given in the hugging face platform for the model is based on the dataset used to train it. Verifying the accuracy of the model for different kinds of images outside the training dataset. The custom dataset mainly includes real and AI generated images related to retail stores generated by GPT4 and midjourney.