no code implementations • 20 Nov 2023 • Beilin Chu, Xuan Xu, Weike You, Linna Zhou
While current image-level detection method fails to generalize to unseen domains, owing to the domain-shift phenomenon brought by CNN's strong inductive bias towards Deepfake texture, video-level one shows its potential to have both generalization across multiple domains and robustness to compression.
no code implementations • 3 Apr 2023 • Xuan Xu, Saarthak Kapse, Rajarsi Gupta, Prateek Prasanna
This marks the first time that ViT has been introduced to diffusion autoencoders in computational pathology, allowing the model to better capture the complex and intricate details of histopathology images.
no code implementations • ICCV 2023 • Aishik Konwer, Xiaoling Hu, Joseph Bae, Xuan Xu, Chao Chen, Prateek Prasanna
We propose a novel approach to learn enhanced modality-agnostic representations by employing a meta-learning strategy in training, even when only limited full modality samples are available.
no code implementations • CVPR 2022 • Aishik Konwer, Xuan Xu, Joseph Bae, Chao Chen, Prateek Prasanna
In our method, a self-attention based Temporal Convolutional Network (TCN) is used to learn a representation that is most reflective of the disease trajectory.
no code implementations • 3 Feb 2022 • Xuan Xu, Prateek Prasanna
Image-based brain cancer prediction models, based on radiomics, quantify the radiologic phenotype from magnetic resonance imaging (MRI).
no code implementations • 2 Dec 2021 • Majid Jaberi-Douraki, Soudabeh Taghian Dinani, Nuwan Indika Millagaha Gedara, Xuan Xu, Emily Richards, Fiona Maunsell, Nader Zad, Lisa Ann Tell
Active ingredients, sensitivity, matrices, and species that have been evaluated for commercial rapid assay tests are typically reported on manufacturers' websites or in PDF documents that are available to consumers but may require a special access request.
no code implementations • 1 Oct 2020 • Xuan Xu, Xin Xiong, Jinge Wang, Xin Li
Thanks to newly designed Deformable Kernel Convolution Alignment (DKC_Align) and Deformable Kernel Spatial Attention (DKSA) modules, DKSAN can better exploit both spatial and temporal redundancies to facilitate the information propagation across different layers.
no code implementations • 14 Sep 2020 • Dario Fuoli, Zhiwu Huang, Shuhang Gu, Radu Timofte, Arnau Raventos, Aryan Esfandiari, Salah Karout, Xuan Xu, Xin Li, Xin Xiong, Jinge Wang, Pablo Navarrete Michelini, Wen-Hao Zhang, Dongyang Zhang, Hanwei Zhu, Dan Xia, Haoyu Chen, Jinjin Gu, Zhi Zhang, Tongtong Zhao, Shanshan Zhao, Kazutoshi Akita, Norimichi Ukita, Hrishikesh P. S, Densen Puthussery, Jiji C. V
Missing information can be restored well in this region, especially in HR videos, where the high-frequency content mostly consists of texture details.
no code implementations • 8 Nov 2019 • Xuan Xu, Yanfang Ye, Xin Li
Image demosaicing and super-resolution are two important tasks in color imaging pipeline.