Machine Intelligence with Deep Learning (Wintersemester 2018/2019)

Lecturer: Dr. Haojin Yang (Internet-Technologien und -Systeme)
Tutors: Joseph Bethge

General Information

• Weekly Hours: 4
• Credits: 6
• Graded: yes
• Enrolment Deadline: 26.10.2018
• Teaching Form: Seminar / Project
• Enrolment Type: Compulsory Elective Module
• Course Language: English
• Maximum number of participants: 20

Programs, Module Groups & Modules

Description

Artificial intelligence (AI) is the intelligence exhibited by computer. This term is applied when a machine mimics "cognitive" functions that humans associate with other human minds, such as "learning" and "problem solving". Currently researchers and developers in this field are making efforts to AI and machine learning algorithms which intend to train the computer to mimic some human skills such as "reading", "listening", "writing" and "making inference" etc. From the year 2006 "Deep Learning" (DL) has attracted more and more attentions in both academia and industry. Deep learning or deep neural networks is a branch of machine learning based on a set of algorithms that attempt to learn representations of data and model their high level abstractions. In a deep network, there are multiple so-called "neural layers" between the input and output. The algorithm is allowed to use those layers to learn higher abstraction, composed of multiple linear and non-linear transformations. Recently DL gives us break-record results in many novel areas as e.g., beating human in strategic game systems like Go (Google’s AlphaGo), self-driving cars, achieving dermatologist-level classification of skin cancer etc. In our current research we focus on video analysis and multimedia information retrieval (MIR) by using Deep-Learning techniques.

Course language: German and English

Topics in this seminar:

• Text Localization with Deep Reinforcement Learning AI systems based on reinforcement learning mostly model the behaviour and interactions of an agent with respect to a well defined outside problem. Popular Examples of recent breakthroughs with deep reinforcement learning are the breakthroughs of Google's AlphaGo system or the recent advances in playing Star Craft with a deep model. In this seminar topic we want to use deep reinforcement learning for the localization of text in images. First, we will learn about the basics of reinforcement learning. Second, we will have a look at Q-learning and how it can be applied to create a deep model with reinforcement learning. Last, we will try to use our acquired knowledge to create an agent that localizes text with deep reinforcement learning.
  Further resources:
  • https://github.com/mpatacchiola/dissecting-reinforcement-learning
  • https://mpatacchiola.github.io/blog/2016/12/09/dissecting-reinforcement-learning.html
  • Active Object Localization with Deep Reinforcement Learning (ICCV'15 paper) 
• Binary Neural Networks Convolutional neural networks have achieved astonishing results in different application areas. Various methods which allow us to use these models on mobile and embedded devices have been proposed. Especially Binary Neural Networks (BNNs) seem to be a promising approach for devices with low computational power or applications which have real-time requirements. In this topic you are going to develop an application, which utilizes BNNs, to be able to run a machine learning model independent of a processing server or a network connection on a low-powered device, e.g. a smartphone or a RaspberryPi. We have possible application ideas prepared, but would like to encourage you to think about your own ideas and will discuss the specific application together at the beginning of the project.
• Interpretable Deep Models Deep Neural Networks (DNNs) have achieved state-of-the-art accuracy in a wide range of practical applications, such as machine translation, object segmentation, speech recognition. However, DNNs are generally considered as black boxes, because given their nonlinearity and deeply nested structure it is difficult to intuitively and quantitatively understand their inference, e.g. what made the trained DNN model arrive at a particular decision for a given data point. Several methods have been developed to understand what a DNN has learned. Some of this work is dedicated to visualize particular neurons or neuron layers, other work focuses on methods which visualize the impact of particular regions of a given input image. An important question for the practitioner is how to objectively measure the quality of an explanation of the DNN prediction and how to use these explanations for improving the model.
  In this regard we will study recently proposed techniques for interpreting, explaining and visualizing deep models and explore their practical usefulness in computer vision.

Requirements

• Strong interests in video/image processing, machine learning (Deep Learning) and/or computer vision

• Software development in C/C++ or Python

• Experience with OpenCV and machine learning applications as a plus

Literature

• Ian J. Goodfellow and Yoshua Bengio and Aaron Courville, "Deep Learning", online version
• Pedro Domingos “The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World”
• Christopher M. Bishop “Pattern Recognition and Machine Learning” google it
• cs231n tutorials: Convolutional Neural Networks for Visual Recognition
• Deep Learning courses at Coursera
• Deep Learning - The Straight Dope deep learning tutorials created by MXNet team
• Caffe: Deep learning framework by the BVLC
• Chainer: A flexible framework of neural networks
• MXNet: A Flexible and Efficient Library for Deep Learning
• Tensorflow: An open-source machine learning framework

Examination

The final evaluation will be based on:

• Initial implementation / idea presentation, 10%

• Final presentation, 20%

• Report/Documentation, 12-18 pages, 30%

• Implementation, 40%

• Participation in the seminar (bonus points)

Dates

Montag, 15.15-16.45

Room H-E.51

15.10.2018 15:15-16:45	Vorstellung der Seminar Themen (PDF)
19 - 22.10.2018	Wahl der Themen  (Anmelden on Doodle)
23.10.2018	Bekanntgabe der Themen- und Gruppenzuordnung
wöchentlich	Individuelle Meetings mit dem Betreuer
03.12.2018	Technologievorträge und geführte Diskussion (je 15+5min)
04.02.2019	Präsentation der Endergebnisse (je 15+5min)
bis Ende Februar 2019	Abgabe von Implementierung und Dokumentation (Latex template)
bis Ende März 2019	Bewertung der Leistungen

Zurück