Francesco Locatello: Learning disentangled representations
Event details
| Date | 23.10.2019 |
| Hour | 14:15 › 15:00 |
| Speaker | Francesco Locatello |
| Location | |
| Category | Conferences - Seminars |
Abstract:
The key idea behind the unsupervised learning of disentangled representations is that real-world data is generated by few explanatory factors of variation (e.g. content + position of objects in an image) which can be recovered by unsupervised learning algorithms.
In this talk, I will discuss the recent progress in the field and challenge some common assumptions. I will discuss the role of inductive biases in the theoretical impossibility of unsupervised learning of disentangled representations and provide a sober look at the performances of state-of-the-art approaches. I will further address the question of how to go beyond purely unsupervised disentanglement in both theory and practice and discuss applications to fairness and visual abstract reasoning.
Bio:
Francesco Locatello is a Doctoral Fellow at the Max Planck ETH Center for Learning Systems supervised by Gunnar Rätsch and Bernhard Schölkopf. He is also working as a research consultant at Google Brain in Zürich. He holds a Google PhD Fellowship in Machine Learning and his work on disentangled representations received a Best Paper Award at ICML 2019.
The key idea behind the unsupervised learning of disentangled representations is that real-world data is generated by few explanatory factors of variation (e.g. content + position of objects in an image) which can be recovered by unsupervised learning algorithms.
In this talk, I will discuss the recent progress in the field and challenge some common assumptions. I will discuss the role of inductive biases in the theoretical impossibility of unsupervised learning of disentangled representations and provide a sober look at the performances of state-of-the-art approaches. I will further address the question of how to go beyond purely unsupervised disentanglement in both theory and practice and discuss applications to fairness and visual abstract reasoning.
Bio:
Francesco Locatello is a Doctoral Fellow at the Max Planck ETH Center for Learning Systems supervised by Gunnar Rätsch and Bernhard Schölkopf. He is also working as a research consultant at Google Brain in Zürich. He holds a Google PhD Fellowship in Machine Learning and his work on disentangled representations received a Best Paper Award at ICML 2019.
Practical information
- General public
- Free
Organizer
- Martin Jaggi