Publications

Commutative Lie Group VAE for Disentanglement Learning

By Xinqi Zhu, Chang Xu, Dacheng Tao
Published in International Conference on Machine Learning (Oral/Long Talk), 2021

We view disentanglement learning as discovering an underlying structure that equivariantly reflects the factorized variations shown in data. We propose to encode the data variations with groups, a structure not only can equivariantly represent variations, but can also be adaptively optimized to preserve the properties of data variations.

Where and What? Examining Interpretable Disentangled Representations

By Xinqi Zhu, Chang Xu, Dacheng Tao
Published in Conference on Computer Vision and Pattern Recognition (Oral, Best Paper Candidate), 2021

Unlike the independence assumption, interpretability has rarely been exploited to encourage disentanglement in the unsupervised setting. In this paper, we examine the interpretability of disentangled representations by investigating two questions: where to be interpreted and what to be interpreted?

Learning Disentangled Representations with Latent Variation Predictability

By Xinqi Zhu, Chang Xu, Dacheng Tao
Published in European Conference on Computer Vision, 2020

This paper defines the variation predictability of latent disentangled representations. Given image pairs generated by latent codes varying in a single dimension, this varied dimension could be closely correlated with these image pairs if the representation is well disentangled.

Approximated Bilinear Modules for Temporal Modeling

By Xinqi Zhu, Chang Xu, Langwen Hui, Cewu Lu, Dacheng Tao
Published in International Conference on Computer Vision, 2019

We consider two less-emphasized temporal properties of video: 1. Temporal cues are fine-grained; 2. Temporal modeling needs reasoning. To tackle both problems at once, we exploit approximated bilinear modules (ABMs) for temporal modeling.

B-CNN: Branch Convolutional Neural Network for Hierarchical Classification

By Xinqi Zhu, Michael Bain
Published in arXiv pre-print, 2017

In this paper, we introduce a variant of the traditional CNN model named the Branch Convolutional Neural Network (B-CNN). A B-CNN model outputs multiple predictions ordered from coarse to fine along the concatenated convolutional layers corresponding to the hierarchical structure of the target classes, which can be regarded as a form of prior knowledge on the output.