[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-120107-en":3,"doc-seo-120107-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":11,"language":21,"language_code":22,"site_id":23,"html_lang":22,"table_of_contents":24,"faqs":25,"seo_title":26,"seo_description":14,"update_tm":27,"read_time":28},120107,8796095462418,"Noah","https://ap-avatar.wpscdn.com/avatar/80000253c1241d02b47?x-image-process=image/resize,m_fixed,w_180,h_180&k=1778826106357471780",6,"Technology","Harnessing Machine Learning for Discerning AI-Generated Synthetic Images - Final Project","AI-generated synthetic images are increasingly difficult to distinguish from authentic visual content, undermining digital media credibility and enabling disinformation and fraud. This project applies machine learning to detect AI-generated images by training deep models on the CIFAKE dataset labeled as “Real” and “Fake”. Transfer learning refines ResNet, VGGNet, and DenseNet for higher identification precision, and results are benchmarked against a vanilla SVM baseline and a custom CNN. Experiments show optimized deep architectures outperform traditional methods, with DenseNet reaching 97.74% accuracy.","Harnessing Machine Learning for Discerning AI-Generated Synthetic Images  \nFinal Project - CS 5785 Applied Machine Learning (2023FA)  \nYuyang Wang Cornell University, USA {[yw2545}@cornell.edu](yw2545}@cornell.edu)  \nYizhi Hao Cornell University, USA {[yz2222}@cornell.edu](yz2222}@cornell.edu)  \nAmando Xu Cong  \nCornell University, USA {[ax45}@cornell.edu](ax45}@cornell.edu)  \narXiv :2401 .07358v2 [ cs .CV] 23 May 2024  \nAbstract—In the realm of digital media, the advent of AIgenerated synthetic images has introduced significant challenges in distinguishing between real and fabricated visual content. These images, often indistinguishable from authentic ones, pose a threat to the credibility of digital media, with potential implications for disinformation and fraud. Our research addresses this challenge by employing machine learning techniques to discern between AI-generated and genuine images. Central to our approach is the CIFAKE dataset, a comprehensive collection of images labeled as “Real” and “Fake”. We refine and adapt advanced deep learning architectures like ResNet, VGGNet, and DenseNet, utilizing transfer learning to enhance their precision in identifying synthetic images. We also compare these with a baseline model comprising a vanilla Support Vector Machine (SVM) and a custom Convolutional Neural Network (CNN). The experimental results were significant, demonstrating that our optimized deep learning models outperform traditional methods, with DenseNet achieving an accuracy of 97.74% . Our application study contributes by applying and optimizing these advanced models for synthetic image detection, conducting a comparative analysis using various metrics, and demonstrating their superior capability in identifying AI-generated images over traditional machine learning techniques. This research not only advances the field of digital media integrity but also sets a foundation for future explorations into the ethical and technical dimensions of AI-generated content in digital media.  \nI. INTRODUCTION AND MOTIVATION  \nThe proliferation of AI-generated synthetic images poses a significant challenge, blurring the lines between reality and digital fabrication. These images, such as the one shown in Figure 1, often indistinguishable from authentic ones, threaten the credibility of digital content and could potentially be exploited for disinformation and fraud. Addressing this issue is not just a technological challenge but a critical step towards preserving the integrity of digital media. Our study is an application project aiming to employ machine learning techniques to efficiently differentiate between AI-generated and genuine images.  \nCentral to our methodology is the CIFAKE dataset [1], exemplified in Figure 2, a comprehensive collection of images categorized as “Real” and “Fake”. This dataset is used for training and testing our models. Our approach includes refining advanced deep learning architectures such as ResNet, VGGNet, and DenseNet through transfer learning. Each of these models, renowned for their effectiveness in image classification tasks, is adapted through transfer learning to suit our specific challenge. This process allows us to capitalize on the strengths of these  \nFigure 1 . AI-generated Image of Pope Francis  \npre-trained models, enhancing their precision and efficiency in identifying synthetic images. Additionally, we trained a vanilla method of Support Vector Machine (SVM) and designed a custom Convolutional Neural Network (CNN) to serve as baseline comparisons. These methods allow us to benchmark the efficiency and accuracy of our models in distinguishing synthetic images.  \nOur study is distinguished by several key contributions:  \n• Application and optimization of advanced deep learning architectures (ResNet, VGGNet, DenseNet) for synthetic image detection.  \n• Development and training of a vanilla SVM and a custom CNN model as baseline methodologies.  \n• Conducting a comparative analysis of our mod","cbCail6JHzZ31myA","https://ap.wps.com/l/cbCail6JHzZ31myA","pdf",4750301,1,"English","en",105,"# Introduction and Motivation\n## Background\n### Importance of Distinguishing Real from Fake\n### Evolution of Machine Learning in Image Analysis\n### Emergence of Deep Learning and CNNs\n### Advancements in Deep Learning","[{\"question\":\"What problem does the project address?\",\"answer\":\"The project targets the difficulty of distinguishing real images from AI-generated synthetic images, which can harm trust and enable disinformation or fraud.\"},{\"question\":\"What dataset is central to the proposed method?\",\"answer\":\"The approach uses the CIFAKE dataset, containing images labeled as “Real” and “Fake” for training and testing.\"},{\"question\":\"Which models are compared, and what is the key result?\",\"answer\":\"ResNet, VGGNet, and DenseNet are built with transfer learning and compared against a vanilla SVM and a custom CNN baseline; DenseNet achieves 97.74% accuracy and outperforms traditional methods.\"}]","Harnessing Machine Learning for Discerning AI-Generated Synthetic Images - Final Project | PDF",1785728232,15,{"code":4,"msg":30,"data":31},"ok",{"site_id":23,"language":22,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"harnessing-machine-learning-for-discerning-ai-generated-synthetic-images-final-project","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/technology/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/harnessing-machine-learning-for-discerning-ai-generated-synthetic-images-final-project/120107/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":22,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does the project address?","Question",{"text":74,"@type":75},"The project targets the difficulty of distinguishing real images from AI-generated synthetic images, which can harm trust and enable disinformation or fraud.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"What dataset is central to the proposed method?",{"text":79,"@type":75},"The approach uses the CIFAKE dataset, containing images labeled as “Real” and “Fake” for training and testing.",{"name":81,"@type":72,"acceptedAnswer":82},"Which models are compared, and what is the key result?",{"text":83,"@type":75},"ResNet, VGGNet, and DenseNet are built with transfer learning and compared against a vanilla SVM and a custom CNN baseline; DenseNet achieves 97.74% accuracy and outperforms traditional methods.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":23},{"code":4,"msg":5,"data":91},[92,96,100,104,109,112,117,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":110,"slug":111},50,"technology",{"id":113,"doc_module":4,"doc_module_name":45,"category_name":114,"show_sort_weight":115,"slug":116},7,"Healthcare",40,"healthcare",{"id":118,"doc_module":4,"doc_module_name":45,"category_name":119,"show_sort_weight":120,"slug":121},8,"Research & Report",30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":105,"slug":137},19,"General","general"]