演講公告

新聞標題：　( 2023-03-08 )

• 演講主題：Generative AI: From Text to Images

• 主講人：蔡炎龍教授 (政治大學應數系)

• 演講日期：2023年3月14日(二) 14:00 – 15:00

• 演講地點：(光復校區) 科學一館223室

• 摘要內容：
  

  Abstract
  
  Generative AI has been a hot topic in recent years, with applications ranging from art and entertainment to healthcare and education. In this talk, we will explore the principles and techniques behind generative models, with a focus on text and image generation.
  
  We will begin by discussing Recurrent Neural Networks (RNNs) and their role in generating text. We will then introduce the concept of transformers and their use in language modeling, leading to a discussion on the GPT series and ChatGPT.
  
  Moving on to image generation, we will discuss the principles of autoencoders and Variational Autoencoders (VAEs), before delving into Generative Adversarial Networks (GANs) and their use in generating realistic images.
  
  Finally, we will introduce diffusion models, which have gained attention in recent years for their ability to generate high-quality samples from complex distributions.
  
  Throughout the talk, we will highlight the key ideas behind each model, its strengths and weaknesses, and real-world applications. We hope this talk will inspire you to explore the exciting world of generative AI and its potential impact on our lives.

• 相關檔案：Talk_1120314.pdf

go back