Most of previous semi-supervised methods that seek to obtain disentangled representations
using variational autoencoders divide the latent representation into two components: the non…

When the three-phase LCL-type converter is attached to a weak grid, its grid current control
will interact with a phase-locked loop (PLL), via the point of common coupling voltage. And, …

Facial expression generation is one of the most challenging and long-sought aspects of
character animation, with many interesting applications. The challenging task, traditionally …

Neural network-based classification methods are often criticized for their lack of interpretability
and explainability. By highlighting the regions of the input image that contribute the most …

Two recent works have shown the benefit of modeling both high-level factors and their related
features to learn disentangled representations with variational autoencoders (VAE). We …

A new ɛ-Keggin polyoxometalate based metal-organic framework, [TBA] 3 [H 4 PMo 8 V Mo
4 VI O 40 Zn 4 ](HTPT) 2 (1) (H 3 TPT = 1,1′:4′,1″-terphenyl-3,4″,5-tricarboxylic acid, …

For LCL-type grid-connected converter, the grid current control interacts with phase-locked
loop (PLL) under weak grid, via point of common coupling (PCC) voltage. And PLL …

An efficient method for the preparation of unsymmetrical ethers from alcohols catalyzed by
sodium bisulfite is reported. The procedure enables the direct dehydration of primary, …

It is necessary to insert a dead-time in sinusoidal pulse width modulation (SPWM) voltage
source inverter (VSI) to prevent a short circuit in the DC link. However, this will cause the dead-…

We present a generative model for the 3D facial expression mesh sequences, from onset to
the termination of a desired expression. We tailor a Transformer VAE architecture: The …