Domain Adaptation

Domain Adaptation

Has deep learning solved all our vision problems?

Domain Adaptation

Domain Adaptation

Different Domains | Same Task

Domain Adaptation

Source data

Target data

Domain Adaptation (DA)

Scenarios:

• Homogeneous DA:                                  Heterogeneous DA:

• Supervised, Semi-Supervised, or Unsupervised

• One-step or Multi-step DA

Domain Adaptation (DA)

Scenarios:

• Homogeneous DA:                                  Heterogeneous DA:

• Supervised, Semi-Supervised, or Unsupervised

• One-step or Multi-step DA

Today's approach: Adversarial-based Domain Adaptation

A Review on GANs & Adversarial Training

A Review on GANs and Adversarial Training

Generator Functions:

A generator can map a known distribution to the distribution on the feature space:

​

A Review on GANs and Adversarial Training

Adversarial Training: 

Design a game between machines where the equilibrium solves a learning problem.

GANs:

• Generator:
• Discriminator: ​
• Game:

A Review on GANs and Adversarial Training

Adversarial Domain Adaptation

Ref: Adversarial Discriminative Domain Adaptation - Tzeng et al. 2017

Adversarial Domain Adaptation

Idea: Consider a intermediate feature space, a common representation for both domains.

Train a classifier on the labeled source data, passing through the representation space

Play a game between the maps and a discriminator

Adversarial Domain Adaptation

Pre-training:

Adversarial Domain Adaptation

Adversarial Adaptation (First turn)

Adversarial Domain Adaptation

Adversarial Adaptation (Second turn)