"Devs are done in 6 months." "Agents will replace whole teams." Stop. The problem is not the tool, it's the vocabulary. Say "agent" and people picture a coworker. What you actually have is an LLM in an execution loop with tool calling. Nothing more.

Positions remain open! Both PhD and postdoctoral opportunities are available on scientific foundation models. An additional position is also available on AI for regional climate models (jointly with @xavierfettweis.bsky.social). Do not hesitate to apply!

Gilles Louppe@glouppe.bsky.social · 2y ago

📣 Hiring! I am looking for PhD/postdoc candidates to work on foundation models for science at @ULiege, with a special focus on weather and climate systems. 🌏 Three positions are open around deep learning, physics-informed FMs and inverse problems with FMs.

The two postdoctoral positions remain open! (the PhD position has been filled) Applications from deep learning researchers or atmospheric scientists are welcome! Ping me if you have any questions!

Gilles Louppe@glouppe.bsky.social · 2y ago

📣 Hiring! I am looking for PhD/postdoc candidates to work on foundation models for science at @ULiege, with a special focus on weather and climate systems. 🌏 Three positions are open around deep learning, physics-informed FMs and inverse problems with FMs.

Where are the women in AI? www.interface-eu.org/publications...

AI's Missing Link: The Gender Gap in the Talent Pool

Our analysis of data on nearly 1.6 million AI professionals worldwide reveals stark gender imbalances. Women comprise only 22% of AI talent globally, with even lower representation at senior levels – ...

interface-eu.org

The serious research we all need: arxiv.org/abs/2501.00536

Phase behavior of Cacio and Pepe sauce

"Pasta alla Cacio e pepe" is a traditional Italian dish made with pasta, pecorino cheese, and pepper. Despite its simple ingredient list, achieving the perfect texture and creaminess of the sauce can ...

arxiv.org

Our newest SoccerNet challenge on Monocular Depth Estimation is up and running! 🥳 Our world is 3D, football games are too, do you have what it takes to lift football analysis from 2D to 3D? 🏟️⚽️

Karen Sanchez@karensanchez.bsky.social · 2y ago

🚨 The wait is over! 4⃣🚀The 4th and final challenge of #SoccerNet2025 is here: “Monocular Depth Estimation”! 📹 Generate precise depth maps using only a single monocular camera and redefine soccer analysis! With this, all four 2025 challenges are officially announced! 🧵 @cvprconference.bsky.social

I wish I stumbled on this pretty and clear intro to CUDA and GPUs years ago : modal.com/gpu-glossary

2⃣🚀 #SoccerNet2025 Let's present our Challenge #2: “Game State Reconstruction” 🏟 Reconstruct a schematic state of the game over time, capturing key player positions and overall game dynamics! 🔗 soccer-net.org/challenges/2... #AIChallenge #SportsAnalytics @cvprconference.bsky.social

At #SIGGRAPHAsia, I just saw a paper talk where many of the examples shown are about making videos of women in short skirts and gym clothes dance to whatever sequence you want from a single image. How does this get published?

Most SISR methods lately: "This new method boasts a jaw-dropping 0.02 PSNR gain and was validated on, wait for it, 19 images. Truly the stuff of legends. 😂 #ScienceIsAwesome #InnovateOrDie"

Our paper "𝟑𝐃 𝐂𝐨𝐧𝐯𝐞𝐱 𝐒𝐩𝐥𝐚𝐭𝐭𝐢𝐧𝐠: 𝐑𝐚𝐝𝐢𝐚𝐧𝐜𝐞 𝐅𝐢𝐞𝐥𝐝 𝐑𝐞𝐧𝐝𝐞𝐫𝐢𝐧𝐠 𝐰𝐢𝐭𝐡 𝟑𝐃 𝐒𝐦𝐨𝐨𝐭𝐡 𝐂𝐨𝐧𝐯𝐞𝐱𝐞𝐬” is now available on Arxiv! 💻Project page for some cool visualizations: convexsplatting.github.io 📓 Paper: arxiv.org/abs/2411.14974

Check out this BEAUTIFUL interactive blog about cameras and lenses ciechanow.ski/cameras-and-...

🚀 🌟 #SoccerNet Challenges 2025 are here! In this 5th edition, we’re excited to unveil four diverse tasks! ⚽️ Check the deadline, there’s plenty of time to craft a groundbreaking solution and rise to the top of the leaderboard! 🏆 🔥Let the Challenges begin! #CVPR2025

SPARS3R: Semantic Prior Alignment and Regularization for Sparse 3D Reconstruction Yutao Tang, Yuxiang Guo, Deming Li, Cheng Peng tl;dr: align Mast3r to colmap point cloud, align outliers separately, as per semantic group -> splat arxiv.org/abs/2411.12592

DGTR: Distributed Gaussian Turbo-Reconstruction for Sparse-View Vast Scenes Hao Li, et al. tl;dr: scene->non-overlap regions->sparse images->Splatt3R+global alignment+depth priors->initialization->local GS models->distillation-based model aggregation arxiv.org/pdf/2411.12309

Future Bsky homes for our #computervision conferences! @cvprconference.bsky.social @iccv.bsky.social @eccv.bsky.social @wacvconference.bsky.social Stay tuned!

Image Matching Filtering and Refinement by Planes and Beyond Fabio Bellavia, Zhenjun Zhao, Luca Morelli, Fabio Remondino tl;dr: refine matches by estimating local homographies with special parameterization MiHo and NCC. Improves even LG/LoFTR. Many additional analyses. arxiv.org/abs/2411.09484

MFTIQ: Multi-Flow Tracker with Independent Matching Quality Estimation Jonas Serych, Michal Neoral, Jiri Matas tl;dr: algorithm for robust chaining optical flow in video + learned quality estimation module. RoMA rules, SeaRAFT is second best. arxiv.org/abs/2411.09551