Open flrngel opened 5 years ago
Hey @flrngel ! Thanks for the notes and the corresponding repository. I was trying to understand the statement made by the authors which you cite in Section 2.1(last paragraph). The statement reads " ... by maximising the MI between the encoded signals( which is bounded by the MI between the input signals)... Could you please elaborate on it? Does it mean I(g(x);g(x+t)) < I(x; x+t), if so why ? Sorry for a naive question.
@MacroMayhem Equation (2) and Appendix A will help you to undertand I guess.
https://arxiv.org/abs/1807.03748 ~big fan of Aaron van den Oord~
Abstract
1. Introduction
2. Contrastive Predicting Coding
2.1. Motivation and Intuitions
2.2. Contrastive Predictive Coding
z_t = g_enc(x_t)
in lthe latent space and produces a context latent representation
c_t = gar(z<=t)`it means
2.3. Noise Contrastive Estimation Loss
2.4. Related work
3. Experiments
3.1. Audio
3.2. Vision
model
experiment result
3.3. Reinforcement Learning
4. Conclusions
Notes
best paper of this week