DualContrast: Unsupervised Disentangling of Content and Transformations with Implicit Parameterization

no code implementations • 27 May 2024 • Mostofa Rafid Uddin, Min Xu

We demonstrate that the existing self-supervised methods with data augmentation result in the poor disentanglement of content and transformations in real-world scenarios.

Harmony: A Generic Unsupervised Approach for Disentangling Semantic Content From Parameterized Transformations

no code implementations • CVPR 2022 • Mostofa Rafid Uddin, Gregory Howe, Xiangrui Zeng, Min Xu

Harmony leverages a simple cross-contrastive learning framework with multiple explicitly parameterized latent representations to disentangle content from transformations.

Contrastive Learning Disentanglement

Cryo-shift: Reducing domain shift in cryo-electron subtomograms with unsupervised domain adaptation and randomization

no code implementations • 17 Nov 2021 • Hmrishav Bandyopadhyay, Zihao Deng, Leiting Ding, Sinuo Liu, Mostofa Rafid Uddin, Xiangrui Zeng, Sima Behpour, Min Xu

Cryo-Electron Tomography (cryo-ET) is a 3D imaging technology that enables the visualization of subcellular structures in situ at near-atomic resolution.

Classification domain classification +2