Sign in

Valerio Marsocci

@valeriomarsocci.bsky.social
964 followers 302 following 46 posts

🌏🌱 trying to make Geospatial Foundation Models work Research Fellow at @ESA PhiLab Previously at @KULeuven, @Cnam PhD in Data Science at @Sapienza website: sites.google.com/uniroma1.it/valeri… #AI4EO #GeoAI #SSL4EO

PostsRepliesMedia
Valerio Marsocci @valeriomarsocci.bsky.social · 27/03/2025
#8 Copernicus-FM This paper introduces: a) a new pre-training dataset; b) a new benchmark dataset; c) a GFM, all based on a diverse set of Copernicus data. ⬆️: really appreciate the grid embeddings part ⬇️: some doubts about claims about generalizability arxiv.org/pdf/2503.11849
040
Valerio Marsocci @valeriomarsocci.bsky.social · 17/03/2025
🔥 🎯 Parameter-Efficient Adaptation of Geospatial Foundation Models through Embedding Deflection (#7) New preprint around :) Incorporating inductive biases specific to MSI can enhance the fine-tuning of large Earth observation models, pre-trained on RGB arxiv.org/pdf/2503.09493
020
Valerio Marsocci @valeriomarsocci.bsky.social · 10/03/2025
#6 Lossy Neural Compression for Geospatial Analytics The authors introduce NC and discuss the characteristics of EO and climate data, w.r.t natural images ⬆️: great entry point ⬇️: no baseline exps arxiv.org/pdf/2503.01505
020
Valerio Marsocci @valeriomarsocci.bsky.social · 24/02/2025
#5 Is SSL on Satellite Imagery Better than ImageNet? A Systematic Study with Sentinel-2 This study pretrains two SSL methods on ImageNet and GeoNet. The improvement with GeoNet is minimal. ⬆️ useful to reduce computation? ⬇️ more considerations about the resolutions? arxiv.org/pdf/2502.10669
020
Valerio Marsocci @valeriomarsocci.bsky.social · 14/02/2025
#4 Galileo Galileo is a family of pretrained RS models designed to flexibly process multimodal RS data. It has two loss: one in the pixel space, one in the latent space. ⬆️: multi-modal/temporal/sensor ⬇️: why just using Sentinel data? arxiv.org/pdf/2502.09356
030
Valerio Marsocci @valeriomarsocci.bsky.social · 05/02/2025
#3 GlaViTU This paper presents a novel world-wide dataset and a novel convolutional-transformer, named Glacier-VisionTransformer-U-Net (GlaViTU), for multitemporal and global glacier mapping. ⬆️ relevant task and nice results ⬇️ weak zero-shot transferability? www.nature.com/articles/s41...
nature.com
Globally scalable glacier mapping by deep learning matches expert delineation accuracy - Nature Communications
A deep learning model using open satellite data for scalable, global glacier mapping is developed. This model matches expert-level accuracy, facilitating more reliable glacier monitoring to support cl...
010
Valerio Marsocci @valeriomarsocci.bsky.social · 30/01/2025
#2 Can Location Embeddings Enhance Super-Resolution of Satellite Imagery? It looks like they can :) ⬆️: validating it on a real-world task ⬇️: is it super-resolution or mapping S2 to NAIP? arxiv.org/pdf/2501.15847
152
Valerio Marsocci @valeriomarsocci.bsky.social · 21/01/2025
#1 Diffusion Models for RS This paper provides a comprehensive review of the applications of diffusion models in remote sensing ⬆️ excellent entry point ⬇️ not sure about the statement about the "inherent denoising ability" of diffusion models arxiv.org/abs/2404.08926
030
Valerio Marsocci @valeriomarsocci.bsky.social · 16/01/2025
Hellooooo 🚀 I'll make it swift: I have just started my new position as Internal Research Fellow at European Space Agency - ESA Phi-Lab I am very happy because it looks like a great place where to do research and because I am back in my beloved hometown, Rome 😍
070
Valerio Marsocci @valeriomarsocci.bsky.social · 14/01/2025
Back on social, after a break (can you guess where?) Last year I decided to do a #50paperschallenge I ended up with 43. Still: 🥵 I read more than 50 papers. I just didn't post all 😇 the strategy worked independently of the posted ones For this reason, this year I will do a #40paperschallenge!
240
Valerio Marsocci @valeriomarsocci.bsky.social · 12/12/2024
#41 Beyond Grid Data GNNs open new possibilities for EO, handling irregular, multi-source datasets (e.g. point clouds) for smarter weather forecasts, disaster relief, etc.. ⬆️: excels at non-Euclidean spatial data ⬇️: limited scalability across diverse data (?) arxiv.org/abs/2411.03223
131
Valerio Marsocci @valeriomarsocci.bsky.social · 06/12/2024
🚀🚀🌏 Are geospatial foundation models really impactful? Check it in our new pre-print! Welcome to **PANGAEA: a global and inclusive benchmark for GFMs** arxiv.org/abs/2412.04204 Check also the public GitHub repo (other news/updates soon): github.com/VMarsocci/pa... a short thread 🧵
2104
Valerio Marsocci @valeriomarsocci.bsky.social · 29/11/2024
#39 TCH in African savannas Can global SatML models solve local challenges? This study finds local models outperform global & fine-tuned models for TCH mapping in Africa ⬆️: interesting set of research questions ⬇️: what about "generalist" geospatial foundation models? arxiv.org/pdf/2411.14354
150
Valerio Marsocci @valeriomarsocci.bsky.social · 28/11/2024
#38 SPECIALIZED FOUNDATION MODELS STRUGGLE TO BEAT SUPERVISED BASELINES Specialized FMs in genomics, satellite imaging, and time series, struggle w.r.t. supervised learning pipelines ⬆️: very relevant work ⬇️: just classification, limiting the real-world capabilities* arxiv.org/abs/2411.02796
120
Reposted by Valerio Marsocci
Noa Garcia @noagarciad.bsky.social · 22/11/2024
As I haven't found it out there yet, I made the Women in computer vision started pack. Many more missing, please let me know how is already in bsky to add them! go.bsky.app/BowzivT
114314
Reposted by Valerio Marsocci
Mark A. Hanson @hansonmark.bsky.social · 19/11/2024
The strain on scientific publishing: we set out to characterise the remarkable growth of the scientific literature in the last few years, in spite of declining growth in total scientists. What is going on? direct.mit.edu/qss/article/... A 🧵 1/n #AcademicSky #PhDchat #ScientificPublishing #SciPub
44994558
Valerio Marsocci @valeriomarsocci.bsky.social · 22/11/2024
🚀 A few weeks ago, we launched Pangaea, a benchmark for geospatial (foundation) models! Focused on dense tasks, Pangea offers a repo to test several models on a global, inclusive and multimodal set of downstream tasks w/ @lebellig.bsky.social @yurujia.bsky.social github.com/yurujaja/pan... 👇
github.com
pangaea-bench/pangaea at main · yurujaja/pangaea-bench
Towards Robust Evaluation for Geospatial Foundation Models - yurujaja/pangaea-bench
171
Reposted by Valerio Marsocci
Kiri Carini @kiricarini.com · 21/11/2024
I didn't see this already, so I went through my follow list and made a Starter Pack for Women+ in geo & EO - very broadly. I'm reminded how lopsided the gender divide is still in our field when I get 1 woman follower for every 10 men. It's particularly noticeable right now. go.bsky.app/Ay1iTTe
5221497
Reposted by Valerio Marsocci
Emily Lines @emilyrlines.bsky.social · 20/11/2024
Interested in a #phd in remote sensing of forests? Want to come work with me in Cambridge using drones and ground-based Lidar and photogrammetry? Check out this funding opportunity for the EPSRC Sensor CDT #phd #forests #remotesensing #lidar 🌳🌲🛰️⚡ www.findaphd.com/phds/program...
findaphd.com
FindAPhD : EPSRC Centre for Doctoral Training in Sensor Technologies and Applications in an Uncertain World (Sensor CDT) at University of Cambridge
Apply for a PhD: EPSRC Centre for Doctoral Training in Sensor Technologies and Applications in an Uncertain World (Sensor CDT) at University of Cambridge
15634
Reposted by Valerio Marsocci
Michele Tobias, PhD @micheletobias.bsky.social · 18/11/2024
It's so fun to see so many people here that I used to interact with in other spaces! There's been a lot of updates to the Open Source Geospatial starter pack. Let me know if you want to be added. go.bsky.app/PGYLmPG
337513
Valerio Marsocci @valeriomarsocci.bsky.social · 21/11/2024
"we have trained more than 50 million neural networks, with more than 150 trillion parameters" 🤯🤯🤯🤯 nianticlabs.com/news/largege... Btw interesting
nianticlabs.com
Building a Large Geospatial Model to Achieve Spatial Intelligence
At Niantic, we are pioneering the concept of a Large Geospatial Model that will use large-scale machine learning to understand a scene and connect it to millions of other scenes globally.
030
Reposted by Valerio Marsocci
Thomas @thomasthefrench.eurosky.social · 21/11/2024
I took the opportunity to make a list of those two starter packs :) bsky.app/profile/did:... I already follow too many people and I need lists to organize myself.
232
Valerio Marsocci @valeriomarsocci.bsky.social · 20/11/2024
#37 ALISE ALISE (ALigned SITS Encoder) is a model for processing irregular and unaligned Satellite Image Time Series (SITS) ⬆️: great sparse data and label handling ⬇️: would be great to extend the domains (geographical and sensor-related) arxiv.org/abs/2407.08448
120
Reposted by Valerio Marsocci
Robbi Bishop-Taylor 🛰️🌏 @satellitesci.bsky.social · 19/11/2024
There's now so many awesome #geospatial people here it's hard to keep track! 🤩 Here's a thread of some of my favourite starter packs so far - let me know what I've missed! 🛰️ Earth Observation & remote sensing: go.bsky.app/4PMRhNL 🌏 GIS: go.bsky.app/TJ7qQF6 @milos-makes-maps.bsky.social (1/n)
1516565
Valerio Marsocci @valeriomarsocci.bsky.social · 19/11/2024
great dataset from great people :)
040
Valerio Marsocci @valeriomarsocci.bsky.social · 19/11/2024
#36 TaxaBind TaxaBind introduces a unified embedding space for 6 data types, solving ecological tasks like species mapping & zero-shot classification ⬆️: robust zero-shot classification ⬇️: what if we want to add/test new tasks? arxiv.org/abs/2411.00683
130
Valerio Marsocci @valeriomarsocci.bsky.social · 19/11/2024
#1 paper I posted before this year was SatCLIP #SatCLIP learns the implicit representations of the image features that characterize a specific location what I liked most: geo adaptation and per-continent performance paper: arxiv.org/pdf/2311.171...
140
Valerio Marsocci @valeriomarsocci.bsky.social · 19/11/2024
👀
030
Valerio Marsocci @valeriomarsocci.bsky.social · 18/11/2024
Hi everyone! 🚀 Happy to be here I am Valerio, from Rome (IT), doing a postdoc at KU Leuven I am working on benchmarking and adapting geospatial foundation models for environmental tasks🌏 Hope to connect with Earth Observation and Deep Learning enthusiasts here! ✨ Follow my #50paperschallenge
191