Главная
Study mode:
on
1
​​ - Introduction
2
- When does dataset bias occur?
3
- Implications in the real-world
4
- Dealing with data bias
5
- Adversarial domain alignment
6
- Pixel space alignment
7
- Few-shot pixel alignment
8
- Moving beyond alignment
9
- Enforcing consistency
10
- Summary and conclusion
Description:
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only! Grab it Explore dataset bias and domain adaptation techniques in this 43-minute lecture from MIT's Introduction to Deep Learning course. Delve into the occurrence and real-world implications of dataset bias, and learn strategies to mitigate its effects. Discover adversarial domain alignment, pixel space alignment, and few-shot pixel alignment methods. Examine approaches that move beyond alignment and enforce consistency in machine learning models. Gain valuable insights from Prof. Kate Saenko of the MIT-IBM Watson AI Lab on taming dataset bias to improve the robustness and fairness of deep learning systems.

Taming Dataset Bias via Domain Adaptation

Alexander Amini
Add to list
0:00 / 0:00