Overview
An in-depth introduction to the core concepts of intelligent user-interfaces. The course primarily deals with machine analysis of human non-verbal behavior and its applications to human-computer, human-robot, and computer-mediated human-human interaction. Methods involve machine learning, deep learning and model based optimization.
Announcements
- 09.05.2017
- Project (Kaggle) and ProjectB (Chalearn) are launched.
- 30.03.2017
- Second assignment is launched.
- 16.03.2017
- Azure platform for the exercises is finally online. Check your email for instructions
- 05.03.2017
- First assignment is launched.
- 23.02.2017
- Course website online
Learning Objectives
Students will learn about fundamental aspects of modern intelligent user interfaces. After completing the course students will have acquired theoretical and practical knowledge about the most important problems in machine understanding of human behavior and how to leverage such understanding in the design of intelligent user-facing technologies.
The core competency acquired through this course is a solid foundation in machine learning and deep-learning algorithms to process and interpret human input into computing systems. In particular, students should be able to develop systems that deal with the problem of recognizing people in images, detecting and describing body parts, inferring their spatial configuration, performing action/gesture recognition from still images or image sequences, also considering multi-modal data, among others. Furthermore, students will be able to leverage models of human behavior in optimization based (algorithmic) design of user interfaces.
Schedule
| Wk. | Date | Content | Slides | Extra Material |
|---|---|---|---|---|
| 1 | 24.02. | IntroductionIntroduction to class contents & admin |
slides | |
| 2 | 02.03. |
ML for HCI Pt. ILinear Models |
slides
slides (annotated) |
ipynb ipynb (Azure) |
| 3 | 09.03. |
ML for HCI Pt. IINon-linear SVM & Decision Trees |
slides
slides (annotated) |
|
| 4 | 16.03. |
ML for HCI Pt. IIIEnsemble Methods |
slides
slides (annotated)
pptx (wth videos) |
|
| 5 | 23.03. |
Dynamic inputHMMs / DL |
slides slides (annotated) | hmm-solution hmm-ipynb (Azure) |
| 6 | 30.03. |
Deep LearningBackprop Algorithm |
slides slides (annotated) | |
| 7 | 06.04. |
Deep LearningConvNets |
slides slides (annotated) | |
| 8 | 13.04. |
Deep LearningSequence Modelling |
slides slides (annotated) | |
| 9 | 20.04. |
No Class (Easter) |
||
| 10 | 27.04. |
Case Study: Programm Committee |
||
| 11 | 04.05. |
User ModellingBasics of Perception |
slides | |
| 12 | 11.05. |
Motor system & text inputFitts' law, language & touch models |
slides | |
| 13 | 18.05. |
Computational UI DesignAlgorithmic design of UIs |
slides | keyopt-ipynb solution |
| 14 | 25.05. |
No Class (Ascension Day) |
Exercises
There will be 3 exercises (2 homework assignments and 1 case study) and one multi-week project. The exercises will constitute 40 % of the final grade. Assignments have to be completed individually. It is ok to discuss with your team members but you have to write your own code.
Exercise sheets and solutions will only be accessible from within the ETH network.
| Exercise | Assignment | Results | Due date |
|---|---|---|---|
| Exercise 1 |
Slides
Jupyter Notebooks
Assignment Page
Register at Kaggle: username@subdomain.ethz.ch |
Grades | 22.03.2017 |
| Exercise 2 |
Slides
Code
Assignment Page
Register at Kaggle: username@subdomain.ethz.ch Slides Code |
Grades | 26.04.2017 |
| Exercise 3 | Program Committee Review Form | Grades Reviews | 27.04.2017 |
| Project |
Slides
Project Codes
Project Page (Kaggle)
Register at Kaggle: username@subdomain.ethz.ch ProjectB Codes ProjectB Page (ChaLearn) |
Grades | 16.06.2017 |
Case Study
We will do one in class case study, simulating a program committee meeting. This is a mandatory and graded part of the course requirements.
Exam
The performance assessment is an oral exam conducted during the examiniation session (Jul-Aug). It will constitute 60% of the final grade.