Today Milind Padalkar is presenting his start-up project NirvanaVisor, using mixed reality and computer vision to support people with vision-related problems. #startup #ai #mixedreality #computervision IIT Alessio Del Bue

We present Contrastive Gaussian Clustering, our novel approach to 3D scene segmentation. By combining 3DGS with a 3D feature field, we can generate novel views as in standard 3DGS, and also generate coherent masks localizing every object in the scene.

PAVIS team will be at Seattle for #CVPR2025 2024. Meet p, Gianluca Scarpellini, Stefano, Alessio Del Bue at Equivision (Tue 18/Summit 321), C3DV (Tue 18/Summit 327) and Embodied AI workshops (Tue 18/Summit 428); presenting DiffAssemble poster on Fri 21. youtu.be/jTWcb9izvLo?si…

There's still some time left to submit! green-fomo.github.io/ECCV2024/call.…

We present 6DGS, our new European Conference on Computer Vision #ECCV2026 paper on camera pose estimation using an NVS model! 📸🔬Our approach proposes a novel "radiant Ellicell" for 3DGS to quickly compare the image with the model. Result? Faster, more accurate one-shot pose recovery. 🧵1/2

🚀 Excited to unveil our latest #ECCV2024 paper: "SelfGeo: Self-supervised and Geodesic-consistent Estimation of Keypoints on Deformable Shapes”where we show how to extract keypoints from non-rigid point clouds without supervision 📚✨ 🧵1/4

🔴 Topological Deep Learning go Directed 🔴 👥 You all know that a directed flow of information between pairs of nodes can be modeled with directed edges 👥 👥 But how can we model a directed flow of information between groups of more than two nodes? 🧵1/3

This year at VISART European Conference on Computer Vision #ECCV2026, we have organised a fantastic panel session with Amanda Wasielewski (Amanda Wasielewski) moderating a panel of Olga Russakovsky (Olga Russakovsky), Augustus E. Wendell (Augustus Wendell) and Mathieu Aubry discussing all things #Art and #ComputerVision on Monday!

Great attendance at our FAILED workshop failed-workshop-eccv-2024.github.io Amazing talk of Olga Russakovsky on fairness in visual recognition!

I'm attending the workshops at European Conference on Computer Vision #ECCV2026 #ECCV2024 in Milan. If you're around, let's have a coffee and discuss the mind-blowing research that's being presented!

Our paper "Maps from Motion (MfM): Generating 2D Semantic Maps from Sparse Multi-view Images" was accepted to 3DV 2025. Thanks to Stefano Stuart James and Alessio Del Bue for the great work! Find out more at our project page: matteot90.github.io/MapsFromMotion/

Don't miss the chance to visit our poster №3.44 at #ACCV2024 We introduce a hybrid representation for human avatars: ❌ Textured mesh - no details ❌ Gaussians - use a lot of memory ✅ Hybrid - nice and compact david-svitov.github.io/HAHA_project_p… My other works: david-svitov.github.io

Printed and ready-ish for NeurIPS Conference to present our collaborative work "Re-assembling the past: The RePAIR dataset and benchmark for real world 2D and 3D puzzle solving" If you are in Vancouver 🇨🇦, feel free to reach out! #MachineLearning #Archeology repairproject2021

PAVIS and former members celebrating the coming holidays and a year full of successes! Well done people, you are an amazing team! IIT

Well done Matteo Toso and Alessio Del Bue (Alessio Del Bue) for presenting our work “Maps from Motion (MfM): Generating 2D Semantic Maps from Sparse Multi-view Images” at International Conference on 3D Vision Paper at: stuart-james.com/publications/2… IIT Comp Sci @ Durham Durham University

Hey there, if you’re an expert in 3D graphics, passionate about GenAI; if you’re a full stack product engineer who loves coding, shipping, and teamwork; and if you’re excited to work with the best pixel AI pioneers on world modeling and spatial intelligence, World Labs would

Surprised but grateful to be selected as an outstanding reviewer #CVPR2025. Sadly, I won't be in person to attend, but I'm sure the program will be great! (No bias) #CVPR2025 #Review

New SfM survey😊 arxiv.org/abs/2505.15814