iiinova — where your digital self keeps evolving

MONTI workshop page: sites.google.com/view/monti20... Full slide deck: drive.google.com/file/d/1ciXS... [n/n]

sites.google.com

MONTI 2026

About MONTI at CVPR 2026

000

The key is to find the right proxy tasks and regularization so that the representations can be optimized at test time without using any golden labels. Thanks so much to Miriam Cha, @sarameghanbeery.bsky.social and the whole organizing team for the invitation to present our work. [3/n]

sites.google.com

MONTI 2026

About MONTI at CVPR 2026

131

MMEarth-Bench: Global Model Adaptation via Multimodal Test-Time Training

My core message was that TTO has potential for improving both the resolution of satellite images and semantic representations used for environmental monitoring tasks. SuperF: sjyhne.github.io/superf/ by @sjyhne.bsky.social MMEarth-Bench: mmearth-bench.com by @luciagordon.bsky.social [2/n]

mmearth-bench.com

Global, multimodal environmental benchmarking datasets for pretrained models and a multimodal test-time training framework for global model adaptation

120

Climate AI Nordics @climateainordics.com · 29/06/2026

𝗠𝗼𝗻𝗶𝘁𝗼𝗿𝗶𝗻𝗴 𝘁𝗵𝗲 𝗪𝗼𝗿𝗹𝗱 𝗧𝗵𝗿𝗼𝘂𝗴𝗵 𝗮𝗻 𝗜𝗺𝗽𝗲𝗿𝗳𝗲𝗰𝘁 𝗟𝗲𝗻𝘀 — This theme of the MONTI workshop at #CVPR2026 fit perfectly with our research! My talk on "𝘖𝘱𝘵𝘪𝘮𝘪𝘻𝘪𝘯𝘨 𝘳𝘦𝘱𝘳𝘦𝘴𝘦𝘯𝘵𝘢𝘵𝘪𝘰𝘯𝘴 𝘢𝘵 𝘵𝘦𝘴𝘵 𝘵𝘪𝘮𝘦 — 𝘍𝘳𝘰𝘮 𝘴𝘶𝘱𝘦𝘳-𝘳𝘦𝘴𝘰𝘭𝘶𝘵𝘪𝘰𝘯 𝘵𝘰 𝘴𝘦𝘮𝘢𝘯𝘵𝘪𝘤 𝘧𝘦𝘢𝘵𝘶𝘳𝘦𝘴" covered two projects. @cvprconference.bsky.social @aicentre.dk 👇 [1/n]

131

Reposted by Nico Lang

💚 Thank you all participants for an amazing 2nd Nordic Workshop on AI for Climate -- a summary of the event can be read here: climateainordics.com/news/2026-06... 🇳🇴 See you next year in Oslo for the 3rd workshop in this annual series -- more info to come!

173

Reposted by Nico Lang

FGVC Workshop @fgvcworkshop.bsky.social · 02/06/2026

It’s almost time for the 13th Workshop on Fine-Grained Visual Categorization! Join us for talks from leaders in the field, great papers + a big community! #CVPR 📅 June 3 🕙 9 AM - 5:45 PM 📍 Room 504 Also, a huge shoutout to our sponsor, @voxel51.bsky.social! Thank you for supporting our community!

0104

Reposted by Nico Lang

Climate AI Nordics @climateainordics.com · 13/05/2026

🌍 CAIN × CCAI: Stronger together across scales. Regional collaboration and global action, accelerating ML for the planet with joint initiatives, knowledge exchange, and capacity building. One goal, two networks, meaningful impact. ⚡️ @climatechangeai.bsky.social climateainordics.com/news/2026-05...

climateainordics.com

New Strategic Partnership: Climate AI Nordics and Climate Change AI

Scaling Global Climate Action: Climate AI Nordics Partners Up with Climate Change AI

143

And more talks from the Global Wetland Center and friends! lnkd.in/ebgjSV3J [n/n]

000

EGU26-7566 | Orals | ITS1.10/BG10.6 Machine Learning and Remote Sensing for Monitoring Tree Biomass Christian Igel Thu, 07 May, 14:05–14:25 (CEST), Room -2.62 lnkd.in/eu5RfjFw [5/n]

EGU26-14707 | Orals | BG9.4 A Vertical Vegetation Structure Model of Europe Hui Zhang, Nico Lang, Stefan Oehmcke, Mikolaj Mazurczyk, Martin Brandt, Rasmus Fensholt, Ankit Kariryaa, and Christian Igel Fri, 08 May, 08:45–08:55 (CEST), Room N1 lnkd.in/eKSydZDr [4/n]

100

EGU26-14956 | Orals | NP4.2 Super-Resolving Any Place on Earth - Implicit Neural Representations for Sentinel-2 Time Series Sander Riisøen Jyhne, Christian Igel, Morten Goodwin, Per-Arne Andersen, Serge Belongie, and Nico Lang Thu, 07 May, 16:15–16:25 (CEST), Room -2.15 lnkd.in/eN_d_GcZ [3/n]

100

EGU26-12736 | Orals | ESSI1.11 MMEarth-Bench: Global Environmental Tasks for Multimodal Geospatial Models Lucia Gordon, Serge Belongie, Christian Igel, and Nico Lang Wed, 06 May, 15:10–15:20 (CEST), Room -2.33 lnkd.in/eXdF29hQ [2/n]

100

I am at EGU in Vienna this week, where everything is about Earth and real problems. Looking forward to an inspiring week and learn from the domain experts! 🌍🌐 Our group will be presenting work on multimodal EO benchmarking, super-resolution, and vertical vegetation structure modelling. 👇 [1/n]

Nico Lang @nicolang.bsky.social · 02/04/2026

This PhD position is part of the Global Wetland Center and the PhD student will be supervised by Christian Igel and myself. Join a vibrant research network through the Danish Pioneer Centre for AI @aicentre.dk, ELLIS @ellis.eu, and @climateainordics.com

032

Reposted by Nico Lang

ELLIS unit Jena @ellisunitjena.bsky.social · 18/03/2026

🚀 𝗗𝗮𝘆 𝗧𝗵𝗿𝗲𝗲 𝗮𝘁 𝘁𝗵𝗲 #ELLISWinterSchool2026 Challenges – Group Work were continued, where participants collaborated intensively. Asst. Prof. @nicolang.bsky.social (University of Copenhagen & Climate AI Nordics) gave a panel talk. 🗓 Our programme: meditwin-project.eu/ellis-winter... @ellis.eu

031

Reposted by Nico Lang

Open-Earth-Monitor Project @openearthmonitor.bsky.social · 11/03/2026

📢 Submit your abstract for the final Open-Earth-Monitor Global Workshop! Join us in Barcelona and share your work on: 🌳Forests & Biodiversity, 🌾Soil–Water–Agriculture, 🔬Climate & Health through #EO, #GeoAI & #opendata. 🌐Details and Submission: earthmonitor.org/global-works...

063

Reposted by Nico Lang

FGVC Workshop @fgvcworkshop.bsky.social · 03/03/2026

How does AI interact with culture? We’re thrilled to have Dr. Maria Antoniak (@mariaa.bsky.social) join us at FGVC! Her interdisciplinary work on AI in the humanities brings a new perspective to our workshop: the fine-grained categorization of the intangible parts of culture. See you at #CVPR2026!

0227

Reposted by Nico Lang

Climate AI Nordics @climateainordics.com · 27/02/2026

The Feb 2026 Climate AI Nordics Newsletter is out! 🌍🤖 In this issue: 🦋 The new Open-Insect biodiversity dataset 🌍 Paola Vesco (PRIO) on ML, climate extremes & conflict 📅 Upcoming AI & climate events 💼 Open PhD & ML roles (RISE, KTH, SLU) Read it here: climateainordics.com/newsletter/2...

climateainordics.com

Climate AI Nordics Newsletter, February 2026

053

Reposted by Nico Lang

FGVC Workshop @fgvcworkshop.bsky.social · 27/02/2026

🚨 Updated Deadline Alert! We have extended the deadline for the Proceedings from Feb 27th to Mar 3rd. #CVPR #CV #AI

032

Longer summary posts on the project: 🌎 www.linkedin.com/posts/lucia-... 🌍 www.linkedin.com/posts/n-lang... [n/n]

000

Team: Lucia Gordon @luciagordon.bsky.social , @serge.belongie.com , Christian Igel, Nico Lang @nicolang.bsky.social Thanks to the Pioneer Centre for AI (P1) @aicentre.dk , Global Wetland Center, and ELIAS - European Lighthouse of AI for Sustainability for supporting this project. [4/n]

Second, we propose a test-time training (TTT) approach as an alternative way of making use of multiple modalities at test time. We show that multimodal pretext tasks can serve as a test-time adaptation signal. [3/n]

100

First, we introduce MMEarth-Bench, a benchmark dataset built to work collectively on the major challenges in remote sensing: Learning from limited reference data, robustness to geographic domain shifts, and resolving ambiguities when estimating biophysical quantities from observational data. [2/n]

100

FathomNet @fathomnet.bsky.social · 19/02/2026

Super excited to have this work out, led by our amazing Lucia Gordon (@luciagordon.bsky.social): “MMEarth-Bench: Global Model Adaptation via Multimodal Test-Time Training” Project page: mmearth-bench.com Paper preprint: arxiv.org/abs/2602.06285 Code and dataset: github.com/lgordon99/mm... [1/n]

174

Reposted by Nico Lang

Are you ready for a challenge? For this year’s FathomNet Kaggle Competition, we’re tackling one of the biggest challenges in computer vision for marine imagery: incomplete labels. Dive in today: www.kaggle.com/competitions... 🤖🌊🧪🐙

kaggle.com

FathomNetCLEF2026 @ LifeCLEF & CVPR-FGVC

Positive-Unlabeled Object Detection in Marine Images

295

Nico Lang @nicolang.bsky.social · 15/01/2026

Recordings of our EurIPS AICC workshop are released! Watch the talks by Philippe Ciais, Sagar Vaze, Natalie Iwanycki Ahlstrand, Drew Purves, and Mirela Beloiu Schwenke. Thanks to @climateainordics.com @aicentre.dk and DDSA - Danish Data Science Academy. @euripsconf.bsky.social

youtube.com

EurIPS 2025 - AICC workshop - YouTube

The AICC workshop was held on Saturday, December 6th, at the University of Copenhagen South Campus as a key part of the EurIPS 2025 conference. The primary g...

074

Nico Lang @nicolang.bsky.social · 14/01/2026

Our FGVC13 workshop is coming to #CVPR2026! Call for papers is out: sites.google.com/view/fgvc13 Big thanks to our artists Mehmet Ozgur Turkoglu and Sophie Stalder for creating our new look and mascot! @cvprconference.bsky.social

031

Reposted by Nico Lang

Oisin Mac Aodha @oisinmacaodha.bsky.social · 05/01/2026

We have two open PhD positions at the interface of AI and ecology. Start dates are Sept 2026. We are looking for candidates with a background in AI/CS, Math, Stats, or Physics that are passionate about solving challenging problems in these domains. Application deadline is in two weeks.

11914

Reposted by Nico Lang

Sara Beery @sarameghanbeery.bsky.social · 25/11/2025

Watch the full talk at go.ted.com/sarabeery If you're interested in learning more about the science behind it, we've just released a preprint describing the methodology and the tool we've built: arxiv.org/abs/2511.15656

go.ted.com

How AI is unearthing hidden scientific knowledge

Scientists estimate that 80 percent of life on Earth is still unknown to humanity. But as global temperatures rise, habitats shrink and food and water sources dry up, we're losing these species faster...

142

Reposted by Nico Lang

Sander R. Jyhne @sjyhne.bsky.social · 16/12/2025

Super-stoked to release this project with @nicolang.bsky.social & Team! Check it out at the respective links, and test the demo to super-resolve Sentinel-2 satellite images anywhere on earth 🌍 🔍

011

The team: Sander Riisøen Jyhne (@sjyhne.bsky.social), Christian Igel, Morten Goodwin, Per-Arne Andersen, Serge Belongie (@serge.belongie.com), Nico Lang (@nicolang.bsky.social) Thanks to the Pioneer Centre for AI (@aicentre.dk) and the Global Wetland Center for supporting this project. [n/n]

011

Hence, our method is not prone to hallucinations, as is often the case with generative approaches. We tested SuperF on both ground-level image bursts and satellite imagery. Ultimately, Sander built a demo app to super-resolve Sentinel-2 satellite images for any place on Earth - try it out. [4/n]

100

In this way, SuperF leverages the continuous nature of INRs for both the sub-pixel alignment and signal reconstruction. Our test-time optimization approach operates solely on low-resolution images and does not require any training on high-resolution data. [3/n]

Building on the established idea of utilising sub-pixel shifts from repeated observations, we developed SuperF, an approach based on implicit neural representations (INRs) that jointly aligns multiple low-resolution images while representing the underlying high-resolution signal. [2/n]

100