Normalizing flows

Last updated Sep 5, 2022 Edit Source

Normalizing flow models are generative models, i.e. they infer the underlying probability distribution of an observed dataset. With that distribution we can do a number of interesting things, namely sample new realistic points and query probability densities

• https://github.com/janosh/awesome-normalizing-flows
• Flow-based Deep Generative Models
• Normalizing flow models
• http://akosiorek.github.io/ml/2018/04/03/norm_flows.html

• #TALK Introduction to Normalizing Flows (ECCV2020 Tutorial)

• #CODE Normalizing Flows in JAX
• #CODE NuX - Normalizing Flows using JAX

• #PAPER NICE: Non-linear Independent Components Estimation (Dinh 2015)
• #PAPER Glow: Generative Flow with Invertible 1x1 Convolutions (Kingma 2018)
  • https://openai.com/blog/glow/
  • #CODE https://github.com/openai/glow
• #PAPER #REVIEW Normalizing Flows: An Introduction and Review of Current Methods (Kobyzev 2020)

See AI/Computer Vision/Image-to-image translation