Computer Vision and Deep Learning

Information:

More Information:

• Description: Each student will choose a paper among a set of proposed papers. The student needs to understand the paper, present it in front of the class and write a short report about it.
• Hot topics: all papers are chosen from active research topics. The one chosen for the seminars of the last semester were presented in 2018 at the main conferences of computer vision, or will be presented in 2019.
• Topics: Computer vision, Object classification/detection/tracking, Supervised/Unsupervised learning, Action classification, Pose estimation, Image generation/superresolution, Style transfer
• Methods: Convolutional Neural Network (CNN), [Variational] Auto-encoder, Recursive Neural Network (RNN), Long-Short Term Model (LSTM), Generative Adversarial Network (GAN)
• Registration required due to limited number of participants.
• Course open for Bachelor and Master students.
• Registration open until: 15th of April
• Register by sending [name, master/bachelor, year, matrikalnumber, major, previous contact to computervision/deeplearning and possible topics of interest] at: johannes.haux@iwr.uni-heidelberg.de
• Preliminary meeting and topic assignment: TBA
• More information can be found in the LSF