Jose Dolz
@josedolz_ets
Followers
312
Following
136
Media
42
Statuses
197
Passionate on medical imaging and computer vision. Associate Professor. ETS Montreal @etsmtl
Montréal, Québec
Joined December 2019
Orgulloso de mi pais en estos momentos tan difíciles 🇪🇸 Proud of my country in these extremely hard moments. All my heart is with mi city and all the victims from this tragedy :( 💔
0
0
4
If you are at #ECCV and want to know about calibrating large language-vision model adaptors, come now to our poster (nr 79)
0
0
5
All about Foundation Models today at the last day of summer school #DLMI with Prof @josedolz_ets! #ChatGPT #FoundationModels #AI #LLMs #VLMs
0
1
2
We begin our day with a talk on Weakly supervised #deeplearning , constrained losses and semantic segmentation! #DLMI2024
0
1
3
so true... lol
Paper completed minutes before conference submission deadline.
0
0
3
I am hiring two post-docs to work at the intersection of medical imaging and machine learning (modelling the uncertainty of large language-vision models). If you are interested, drop me an email for more information jose.dolz@etsmtl.ca
0
1
8
If you are interested in doing a PhD in deep learning and computer vision at ETS Montreal, drop me an email with your CV and research interests (more info in the attached image)
0
3
11
Journal -> Conference!
Do you have a fantastic paper almost ready to submit but waiting for the next conference deadline? Papers accepted for publication at MELBA have the option to be presented at the Medical Imaging with Deep Learning Conference, which will be held in Paris next summer.
0
2
3
5/5. Link to the paper: https://t.co/0PiBZ7QIbY Github: https://t.co/pBtEPCkF5N Congrats to all co-authors!! @bing_bingyuan @adrian_galdran @IsmailBenAyed1
github.com
Code for the paper : Do we really need dice? The hidden region-size biases of segmentation losses. MeDIA 2023. https://www.sciencedirect.com/science/article/abs/pii/S136184152300275X - by-liu/SegLo...
1
0
2
4/5. Dice has an intrinsic bias towards specific extremely imbalanced solutions, whereas CE implicitly encourages the ground-truth region proportions. This explains the wide experimental evidence in medical-imaging, where Dice loss brings improvements for imbalanced segmentation.
1
0
1
3/5.And the second one, a region-size penalty term imposing different biases on the size (or proportion) of the predicted regions. Our information-theoretic analysis uncovers hidden region-size biases.
1
0
0
2/5. In this work, we provide a theoretical analysis, which shows that CE and Dice share a deep connection. They both decompose into two components. The first one, a similar ground-truth matching term, which pushes the predicted foreground regions towards the ground-truth;
1
0
2
1/5.Do you wonder which is the best loss function to use in your medical segmentation model? It is widely argued within the medical-imaging community that Dice and CE losses are complementary, which has motivated the use of compound CE-Dice losses (the de-facto solution nowadays)
2
2
16
🚨One of our latest papers, where we propose to use Denoising Auto-Encoders to model the uncertainty of the predictions in semi-supervised segmentation has been accepted in MedIA. Congrats @sukeshadiga !! 🎉🎉💪💪 Arxiv: https://t.co/UQW5ckj2w1 Github:
github.com
Anatomically-aware Uncertainty for Semi-supervised Image Segmentation - adigasu/Anatomically-aware_Uncertainty_for_Semi-supervised_Segmentation
0
2
14
Kudos to all the first authors and co-authors on these papers!! @93Balamuralim @jul_nicol @IsmailBenAyed1 @imtiaz_masud
0
0
1
2) MoP-CLIP: A Mixture of Prompt-Tuned CLIP Models for Domain Incremental Learning. Paper: https://t.co/LOBH2lHJCW
1
0
3
It seems some of my students will go to Hawaii this winter to present their works at @wacv_official 1) Prompting classes: Exploring the Power of Prompt Class Learning in Weakly Supervised Semantic Segmentation. Paper: https://t.co/fs5cUV7tRL Github: https://t.co/lRG9Q7xzAo
1
3
14