Disentangling by Factorising
About
We define and address the problem of unsupervised learning of disentangled representations on data generated from independent factors of variation. We propose FactorVAE, a method that disentangles by encouraging the distribution of representations to be factorial and hence independent across the dimensions. We show that it improves upon $\beta$-VAE by providing a better trade-off between disentanglement and reconstruction quality. Moreover, we highlight the problems of a commonly used disentanglement metric and introduce a new metric that does not suffer from them.
Hyunjik Kim, Andriy Mnih• 2018
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Time Series Forecasting | ETTh1 | MSE0.484 | 836 | |
| Time Series Forecasting | ETTh1 (test) | MSE0.87 | 398 | |
| Time Series Forecasting | ETTm1 | MSE2.326 | 363 | |
| Time Series Forecasting | ETTm1 (test) | MSE2.326 | 315 | |
| Time Series Forecasting | Traffic (test) | MSE0.982 | 272 | |
| Time Series Forecasting | Weather (test) | MSE0.53 | 248 | |
| Time Series Forecasting | Traffic | MSE0.936 | 211 | |
| Time Series Forecasting | Electricity (test) | MSE0.807 | 130 | |
| Image Reconstruction | CelebA-HQ (test) | FID (Reconstruction)87.19 | 50 | |
| Time Series Forecasting | Wind | MSE0.898 | 46 |
Showing 10 of 71 rows
...