Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

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

TaskDatasetResultRank
Time Series ForecastingETTh1
MSE0.484
601
Time Series ForecastingETTm1
MSE2.326
334
Time Series ForecastingETTh1 (test)
MSE0.87
262
Time Series ForecastingETTm1 (test)
MSE2.326
196
Time Series ForecastingTraffic (test)
MSE0.982
192
Time Series ForecastingTraffic
MSE0.936
145
Time Series ForecastingWeather (test)
MSE0.53
110
Time Series ForecastingElectricity (test)
MSE0.807
72
Image ReconstructionCelebA-HQ (test)
FID (Reconstruction)87.19
50
Disentangled Representation LearningCars3D
FactorVAE0.906
35
Showing 10 of 48 rows

Other info

Follow for update