@valeriomarsocci
🌏🌱 trying to make Geospatial Foundation Models work Research Fellow at @ESA PhiLab Previously at @KULeuven, @Cnam PhD in Data Science at @Sapienza website: https://sites.google.com/uniroma1.it/valeriomarsocci #AI4EO #GeoAI #SSL4EO
#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
🔥 🎯 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
#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
#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
#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
#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...
#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
#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
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 😍
uoooo great news, how many are you trying to cover?
I start the new challenge this week :)
Also, other very cool personal news is coming out
So stay tuned if interested ✨
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!
#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
Did we overlook something? Are you interested in this kind of topic?
We are already considering future updates, so feel free to reach out to give feedbacks and to talk about geospatial foundation models
✨🌏
A great team collaborated on it!
Thx @yurujia.bsky.social @lebellig.bsky.social @nshaud.bsky.social and all the others 🤩
🧵
We observed interesting insights, such as:
1. generally speaking GFMs don't really excel when compared to supervised baselines
2. for some specific scenarios (e.g. HR data), it makes sense to use them
3. multi-temporal data are still under-estimated
other insights in the paper!
🧵
With this benchmark (PANGAEA), we tried to address the following research challenges:
* provide a robust evaluation protocol to benchmark GFMs
* investigate GFMs capabilities, with a focus on a) domain generalization, b) comparison to supervised baselines, c) performance with limited labels
🧵
We collected 11 datasets to create an inclusive, diverse benchmarks, based on these criteria:
* application domain
* geographical distribution
* type of task
* modality
* temporality
Spoiler: no patch-level classification tasks are included!
🧵
🚀🚀🌏
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 🧵
Another paper shows that global models are not always the best choice.
If you are interested in this topic, and in geospatial foundation models in general, next week we will publish an interesting pre-print, connected to our Pangaea repo
Check it here: github.com/yurujaja/pan...
#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
also, in the past I posted about this interesting benchmark paper:
#32 GeoFMs for crop type mapping
it investigates the ability of geoFMs to transfer to new geographic regions in agriculture
⬆️the pivotal topic for real-world applications
⬇️the limited number of geoFMs
arxiv.org/pdf/2409.09451
here is the GitHub of PANGAEA code (that we used for the experiments):
github.com/yurujaja/pan...
🚀🌏
If you want to see how geospatial foundation models are working in real-world tasks w.r.t. supervised baselines, stay tuned cause next week we are releasing the pre-print of PANGAEA, showing interesting results on this topic!
#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
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
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
👉next week I will post some papers on this topic to be ready for our preprint release
⁉️if you have suggestions or questions let us know!
🌍🌎🌏
🚀 What’s Next?
A paper detailing Pangaea with many results is coming in 1-2 weeks
It highlights the differences with other benchmarks, and shows interesting insights on models' performance
🤯Spoiler: geospatial foundation models are far from being generalist
👇
🌟 Why is Pangaea relevant?
- Comprehensive Coverage: global datasets from multiple domains
- Multimodal & Multitemporal Data: diverse sensors both single and multi-temporal
- Easy to Extend: you can contribute with models and datasets
👇