392115 Deep Learning for AI (S) (WiSe 2019/2020)

In recent years, the use of deep neural networks as function approximators has enabled researchers to solve increasingly complex problems. The emerging field of deep reinforcement learning [1] has led to remarkable empirical results in rich and varied domains like robotics [2], arcade games [3], and multiagent interaction [13]. This seminar will help students gain a high-level view about the current state of the art and potential directions for future research [5]. Participants are expected to actively participate in the course by presenting selected articles, and discussing code implementation of the deep learning architectures.

Please, select an article for presentation from these lists [15, 16] or from [5-12].

[1] https://deepmind.com/blog/deep-reinforcement-learning
[2] https://blog.openai.com/learning-dexterity
[3] DQN: https://youtu.be/XjsY8-P4WHM?t=48s
[4] https://www.youtube.com/watch?v=-L4tCIGXKBE
[5] https://www.dropbox.com/s/99eyutemrdb17kj/SIAM%202018.pdf
[6] YOLO COCO Object Detection: https://youtu.be/yQwfDxBMtXg
[7] https://github.com/tensorflow/models/tree/master/research
[8] https://github.com/cbfinn/gps
[9] https://github.com/brendenlake/BPL
[10] https://github.com/deepmind/dm_control
[11] https://github.com/atenpas/gpd
[12] ORB-SLAM2: https://www.youtube.com/watch?v=ufvPS5wJAx0
[13] https://www.pommerman.com
[14] https://www.aicrowd.com
[15] Articles for presentation part 1: https://rebrand.ly/DRL
[16] Articles for presentation part 2: https://docs.google.com/spreadsheets/d/1xvail8eldqtcOeKw-XJK9igRM_tmO_5Zlr5hBKfWxoA

Literaturangaben

Articles for presentation part 1: https://rebrand.ly/DRL
Articles for presentation part 2: https://docs.google.com/spreadsheets/d/1xvail8eldqtcOeKw-XJK9igRM_tmO_5Zlr5hBKfWxoA