1. Using Generative Adversial Networks to Create Training Data from only Textual Descriptions Open Access Author: Cao, Jeffery Xufeng Title: Using Generative Adversial Networks to Create Training Data from only Textual Descriptions Area of Honors: Computer Science Keywords: GANmachine learningCNNtext-to-imageImage Classification File: Download JefferyThesisFinal.pdf Thesis Supervisors: Thomas F La Porta, Thesis SupervisorTing He, Thesis Honors Advisor
2. Robust Image Classification based on Pixel Importance Open Access Author: Chun, Chaewan Title: Robust Image Classification based on Pixel Importance Area of Honors: Computer Science Keywords: adversarial attackimage analysismachine learningCNNheatmap-based blurringrobustness File: Download Schreyer_Honors_Thesis_Chun.pdf Thesis Supervisors: Jia Li, Thesis SupervisorDanfeng Zhang, Thesis Honors Advisor