Zobrazeno 1 - 10
of 201
pro vyhledávání: '"Dolz, José"'
Autor:
Shakeri, Fereshteh, Huang, Yunshi, Silva-Rodríguez, Julio, Bahig, Houda, Tang, An, Dolz, Jose, Ayed, Ismail Ben
Integrating image and text data through multi-modal learning has emerged as a new approach in medical imaging research, following its successful deployment in computer vision. While considerable efforts have been dedicated to establishing medical fou
Externí odkaz:
http://arxiv.org/abs/2409.03868
Autor:
Beizaee, Farzad, Lodygensky, Gregory A., Adamson, Chris L., Thompso, Deanne K., Cheon, Jeanie L. Y., Anderso, Alicia J. Spittl. Peter J., Desrosier, Christian, Dolz, Jose
Lack of standardization and various intrinsic parameters for magnetic resonance (MR) image acquisition results in heterogeneous images across different sites and devices, which adversely affects the generalization of deep neural networks. To alleviat
Externí odkaz:
http://arxiv.org/abs/2407.15717
This paper addresses the critical issue of miscalibration in CLIP-based model adaptation, particularly in the challenging scenario of out-of-distribution (OOD) samples, which has been overlooked in the existing literature on CLIP adaptation. We empir
Externí odkaz:
http://arxiv.org/abs/2407.13588
Despite the significant progress in deep learning for dense visual recognition problems, such as semantic segmentation, traditional methods are constrained by fixed class sets. Meanwhile, vision-language foundation models, such as CLIP, have showcase
Externí odkaz:
http://arxiv.org/abs/2404.08181
Autor:
Huang, Yunshi, Shakeri, Fereshteh, Dolz, Jose, Boudiaf, Malik, Bahig, Houda, Ayed, Ismail Ben
In a recent, strongly emergent literature on few-shot CLIP adaptation, Linear Probe (LP) has been often reported as a weak baseline. This has motivated intensive research building convoluted prompt learning or feature adaptation strategies. In this w
Externí odkaz:
http://arxiv.org/abs/2404.02285
State-of-the-art semi-supervised learning (SSL) approaches rely on highly confident predictions to serve as pseudo-labels that guide the training on unlabeled samples. An inherent drawback of this strategy stems from the quality of the uncertainty es
Externí odkaz:
http://arxiv.org/abs/2403.15567
In this work, we present a novel approach to calibrate segmentation networks that considers the inherent challenges posed by different categories and object regions. In particular, we present a formulation that integrates class and region-wise constr
Externí odkaz:
http://arxiv.org/abs/2403.12364
Autor:
Silva-Rodriguez, Julio, Chelbi, Jihed, Kabir, Waziha, Chakor, Hadi, Dolz, Jose, Ayed, Ismail Ben, Kobbi, Riadh
Using deep learning models pre-trained on Imagenet is the traditional solution for medical image classification to deal with data scarcity. Nevertheless, relevant literature supports that this strategy may offer limited gains due to the high dissimil
Externí odkaz:
http://arxiv.org/abs/2401.15526
Autor:
Murugesan, Balamurali, Vasudeva, Sukesh Adiga, Liu, Bingyuan, Lombaert, Hervé, Ayed, Ismail Ben, Dolz, Jose
Ensuring reliable confidence scores from deep neural networks is of paramount significance in critical decision-making systems, particularly in real-world domains such as healthcare. Recent literature on calibrating deep segmentation networks has res
Externí odkaz:
http://arxiv.org/abs/2401.14487
Efficient transfer learning (ETL) is receiving increasing attention to adapt large pre-trained language-vision models on downstream tasks with a few labeled samples. While significant progress has been made, we reveal that state-of-the-art ETL approa
Externí odkaz:
http://arxiv.org/abs/2312.12730