Zobrazeno 1 - 10
of 841
pro vyhledávání: '"Gomez, Alberto"'
We investigate the utility of diffusion generative models to efficiently synthesise datasets that effectively train deep learning models for image analysis. Specifically, we propose novel $\Gamma$-distribution Latent Denoising Diffusion Models (LDMs)
Externí odkaz:
http://arxiv.org/abs/2409.19371
Autor:
Bransby, Kit M., Kim, Woo-jin Cho, Oliveira, Jorge, Thorley, Alex, Beqiri, Arian, Gomez, Alberto, Chartsias, Agisilaos
Building an echocardiography view classifier that maintains performance in real-life cases requires diverse multi-site data, and frequent updates with newly available data to mitigate model drift. Simply fine-tuning on new datasets results in "catast
Externí odkaz:
http://arxiv.org/abs/2407.21577
Autor:
Bransby, Kit Mills, Beqiri, Arian, Kim, Woo-Jin Cho, Oliveira, Jorge, Chartsias, Agisilaos, Gomez, Alberto
Neural networks can learn spurious correlations that lead to the correct prediction in a validation set, but generalise poorly because the predictions are right for the wrong reason. This undesired learning of naive shortcuts (Clever Hans effect) can
Externí odkaz:
http://arxiv.org/abs/2406.19148
Autor:
Reynaud, Hadrien, Meng, Qingjie, Dombrowski, Mischa, Ghosh, Arijit, Day, Thomas, Gomez, Alberto, Leeson, Paul, Kainz, Bernhard
To make medical datasets accessible without sharing sensitive patient information, we introduce a novel end-to-end approach for generative de-identification of dynamic medical imaging data. Until now, generative methods have faced constraints in term
Externí odkaz:
http://arxiv.org/abs/2406.00808
Autor:
Miguel-Gómez, Alberto
We provide a model-theoretic classification of the countable homogeneous $\mathbf{H}_4$-free 3-hypertournament studied by Cherlin, Hubi\v{c}ka, Kone\v{c}n\'y, and Ne\v{s}et\v{r}il. Our main result is that the theory of this structure is $\mathrm{SOP}
Externí odkaz:
http://arxiv.org/abs/2404.04381
U-Net style networks are commonly utilized in unsupervised image registration to predict dense displacement fields, which for high-resolution volumetric image data is a resource-intensive and time-consuming task. To tackle this challenge, we first pr
Externí odkaz:
http://arxiv.org/abs/2307.02997
Autor:
Kerdegari, Hamideh, Phung1, Tran Huy Nhat, Nguyen, Van Hao, Truong, Thi Phuong Thao, Le, Ngoc Minh Thu, Le, Thanh Phuong, Le, Thi Mai Thao, Pisani, Luigi, Denehy, Linda, Consortium, Vital, Razavi, Reza, Thwaites, Louise, Yacoub, Sophie, King, Andrew P., Gomez, Alberto
Skeletal muscle atrophy is a common occurrence in critically ill patients in the intensive care unit (ICU) who spend long periods in bed. Muscle mass must be recovered through physiotherapy before patient discharge and ultrasound imaging is frequentl
Externí odkaz:
http://arxiv.org/abs/2306.04739
We propose a novel pipeline for the generation of synthetic ultrasound images via Denoising Diffusion Probabilistic Models (DDPMs) guided by cardiac semantic label maps. We show that these synthetic images can serve as a viable substitute for real da
Externí odkaz:
http://arxiv.org/abs/2305.05424
Autor:
Reynaud, Hadrien, Qiao, Mengyun, Dombrowski, Mischa, Day, Thomas, Razavi, Reza, Gomez, Alberto, Leeson, Paul, Kainz, Bernhard
Image synthesis is expected to provide value for the translation of machine learning methods into clinical practice. Fundamental problems like model robustness, domain transfer, causal modelling, and operator training become approachable through synt
Externí odkaz:
http://arxiv.org/abs/2303.12644
Inspired by the work of Geiss, Leclerc and Schr\"oer [Represent. Theory 20, (2016)] we realize the enveloping algebra of the positive part of an affine Kac-Moody Lie algebra of Dynkin type $\tilde{\mathsf{C}}_n$ as an convolution algebra of construct
Externí odkaz:
http://arxiv.org/abs/2303.06260