• Yarramreddy Prudhvi
• Thota Adinarayana
• Turla Chandu
• Shaik Musthak
• Govada Sireesha

As artificial intelligence (AI) continues to progress in the creation of realistic content in domains like media, advertising and security, it is essential to address the growing concerns about image authenticity. To address this issue, we propose an innovative image classification model that uses TensorFlow, T4 GPU, and a diverse dataset for rigorous training. The model includes strategic information augmentation and normalization, as well as dataset splitting. The model is based on the robust architecture of ResNet50, which is well-known for deep learning capabilities. The model adds additional layers for better feature extraction and represent learning. The integration of dense units and batch normalization and drop out techniques further improves the model’s effectiveness. Early results demonstrate the promise of this model in distinguishing real from AI-generated image, which will significantly contribute to the ongoing discussion on image authenticity, and provide a strong response to the challenges of synthetic content in today’s technological landscapes.

Deep Learning,ResNet50,Classification, AI, Real

"Deeplearning Approach for Classifying Real and Computer-Generated Images", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 3, page no.e147-e154, March-2024, Available :http://www.jetir.org/papers/JETIR2403420.pdf

"Deeplearning Approach for Classifying Real and Computer-Generated Images", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 3, page no. ppe147-e154, March-2024, Available at : http://www.jetir.org/papers/JETIR2403420.pdf