Pushpak Pati รีทวีตแล้ว
Happy to share that our paper "Differentiable Zooming for Multiple Instance Learning on Whole-Slide Images" has been accepted at #ECCV2022! 🎉
arXiv: arxiv.org/abs/2204.12454
Code: github.com/histocartograp…
🧵1/6
English
Pushpak Pati
36 posts
Pushpak Pati
@pushpak_pati
Senior Scientist @ Johnson & Johnson
Switzerland เข้าร่วม Haziran 2012
193 กำลังติดตาม140 ผู้ติดตาม
Pushpak Pati รีทวีตแล้ว
Vision GNN: An Image is Worth Graph of Nodes
abs: arxiv.org/abs/2206.00272
github: github.com/huawei-noah/CV…
propose to represent the image as a graph structure and introduce a new Vision GNN (ViG) architecture to extract graph level feature for visual tasks
English
Pushpak Pati รีทวีตแล้ว
Proud to share our 150-page "proto-book" with @mmbronstein @joanbruna @TacoCohen on geometric DL! Through the lens of symmetries and invariances, we attempt to distill "all you need to build the architectures that are all you need".
geometricdeeplearning.com
More info below! 🧵
English
Pushpak Pati รีทวีตแล้ว
Motivation Monday of an exciting week. Tomorrow 🕘 9-12 am: Building Interpretable AI for Digital Pathology @appliedmldays. Tickets from 30 CHF. Last chance to register to do some interpretability jazz with me, @pushpak_pati, @GuillaumeJaume and our special guest: Inti Zlobec
GIF
English
Pushpak Pati รีทวีตแล้ว
1/ Want to explore the paradigm of graphs in computational pathology?
Meet histocartography, a one-stop #python library to build, model and interpret entity-graphs!
github.com/histocartograp…
@IBMResearch @pushpak_pati @afoncubiertar @mariagabrani
More info 👇
English
Pushpak Pati รีทวีตแล้ว
On April 27th we look forward to hosting the workshop "Building Interpretable AI for Digital Pathology" led by @mormontre, @pushpak_pati and @GuillaumeJaume. Find out more here buff.ly/32fuwg1
English
4/ Some papers already using BRACS:
- Hierarchical Graph Representations for Digital Pathology, arxiv.org/abs/2102.11057
- Quantifying Explainers of Graph Neural Networks in Computational Pathology, CVPR, 2021, arxiv.org/abs/2011.12646
English
3/ A challenging and realistic dataset that includes tumors of various sizes and aspect ratio, staining variation, tissue preparation artifacts.
English
1/ ⏰ The BRACS dataset is out!
Meet the large cohort of H&E stained breast cancer RoIs.
A joint effort by @istitutotumori, @ICAR_CNR & @IBMResearch
@GuillaumeJaume @afoncubiertar @mariagabrani
More info 👉 bracs.icar.cnr.it & 👇
English
Pushpak Pati รีทวีตแล้ว
1⃣Interpretable AI for Digital Pathology
🗓️27 April 9-12h (CET)
👩💻MSc, PhD students & researchers
💰69 (students) - 120 (rest) (whole event)
🗺️Online
@mormontre @pushpak_pati & @GuillaumeJaume will give you hands-on tutorials to apply interpretability
👉appliedmldays.org/events/amld-ep…
English
Pushpak Pati รีทวีตแล้ว
Here a line up of encouraging activities on Interpretable #AI for these spring days!
Register to appliedmldays.org/events/amld-ep… for attending the1⃣event: Building Interpretable AI for Digital Pathology
Me, @pushpak_pati and @GuillaumeJaume will be giving > 2 hours coding tutorials
MedGIFT@MedGIFT_group
You probably were not expecting this but... Here are the "Interpretable #AI Spring Days" for younger & experienced researchers ‼️ 1⃣Interpretable AI for Digital Pathology 2⃣Towards a Global Taxonomy for Interpretable AI 3⃣Introduction to Interpretable AI Check👇for details 🧵
English
Pushpak Pati รีทวีตแล้ว
Glad to announce that our work on Quantifying Explainers of Graph Neural Networks in Computational Pathology has been accepted to #CVPR2021!
@pushpak_pati @afoncubiertar @mariagabrani @IBMResearch
Guillaume Jaume@GuillaumeJaume
1/ Beyond saliency maps: How to quantify and rank #explainability methods in #digitalpathology? Checkout our new preprint arxiv.org/pdf/2011.12646… based on cell graph analysis. @IBMResearch @pushpak_pati @afoncubiertar @mariagabrani
English
Pushpak Pati รีทวีตแล้ว
Yay! Me, @pushpak_pati and @GuillaumeJaume will be talking about Interpretable AI for Digital Pathology for a half-day workshop at the @appliedmldays. Join us :)
MedGIFT@MedGIFT_group
Join us to hear about visualization methods, concept attribution and interpretable graph-networks ‼️ 📅 April 27 ⏲️9-12h (CET) 💻Online (@appliedmldays) #interpretability #DeepLearning #pathology #GirlsWhoCode #MachineLearning #onlinelearning
English
Pushpak Pati รีทวีตแล้ว
Alongside @pl219_Cambridge, I'll be teaching a Master's module on GNNs @Cambridge_Uni. Feels surreal to be on the (virtual) other side 😊
Based on @williamleif's book + my ongoing work w/ @joanbruna @mmbronstein @TacoCohen. Hope to release materials soon! cl.cam.ac.uk/teaching/2021/…
English
Pushpak Pati รีทวีตแล้ว
New year is a good time to is a good time to recap and make predictions. In a new post in @TDataScience I sought the opinion of 12 prominent researchers in the field of #GraphML to predict what is in store for 2021.
towardsdatascience.com/predictions-an…
English
Pushpak Pati รีทวีตแล้ว
My main talks on Graph Neural Networks in 2020
1. Introduction to GNNs
youtube.com/watch?v=Iiv9R6…
2. Recent developments in GNNs
youtube.com/watch?v=M60hux…
3. Benchmarking GNNs
youtube.com/watch?v=tuChBS…
Hope they can be useful.
Happy new year to everyone !
YouTube
YouTube
YouTube
English
@simongraham73 @TIAwarwick @PathLAKE_CoE @vqdang93 @nmrajpoot @sea_raza @mitkoveta @sakhurram @AI4Pathology @anantm @NormanZerbe @MTalhaQaiser @QiDou_ @jnkath @LawrenceYoung11 Awesome. This will be very useful for future developments in #digitalpathology.
English
[1/3] The PyTorch implementation of HoVer-Net is now available 🎉🎉! Go and check it out at github.com/vqdang/hover_n…
@TIAwarwick @PathLAKE_CoE @vqdang93 @nmrajpoot @sea_raza
English
Pushpak Pati รีทวีตแล้ว
We are excited to share our #NeurIPS2020 papers presented this week from our lab. Check the below teaser to get a glimpse of the topics. See you at the poster sessions!
youtu.be/thNYTHidAKs
YouTube
English
Checkout our new preprint arxiv.org/pdf/2011.12646… for quantifying explainability methods in #DigitalPathology
Guillaume Jaume@GuillaumeJaume
1/ Beyond saliency maps: How to quantify and rank #explainability methods in #digitalpathology? Checkout our new preprint arxiv.org/pdf/2011.12646… based on cell graph analysis. @IBMResearch @pushpak_pati @afoncubiertar @mariagabrani
English