Our paper on onboard ML processing is finally out! 🧠🚀
Here we show that we can fine tune a model and re-upload its weights to the satellite to make new inferences in orbit.
Full article at https://t.co/ku6KSxM7ni 📰
New experiments show that in-orbit ML payloads can be retrained on new images — step forward for Space + ML applications. Read the full article in Nature Scientific Journal: https://t.co/AhVaDyuE8Z
🚨 Flash Flood Map of #Valencia 🌊
We developed new maps using satellite data and machine learning to show the full impact of recent floods. #EO and #AI for disaster response.
https://t.co/bnyzXrOKBQ
#ValenciaFloods#EarthObservation#DANA
El que anava a ser un gran corredor verd al sud de València, l’alcaldessa ha decidit clavar-li cotxes en els dos sentits i travessant, el que podria ser un gran pulmó serà altra autopista urbana.
A��ò sols ho pararem com ja paràrem l’autopista que volien fer al llit del Túria: Al carrer!
😶🌫️ CloudSEN12+ is out! Fantastic (quality and size) dataset for cloud segmentation in Sentinel-2 images! @CopernicusEU
Superb work, Cesar Aybar, Lesly Bautista and the team @isp_uv_es
💽 Data https://t.co/wFAMwCx1QF
📜 Paper https://t.co/hXv6bB1cpl
🗺️🤗 Dataset https://t.co/SzXaNyHo6Q
💻🤗 Trained models https://t.co/nvnROJtexg
📢Step by step tutorials https://t.co/d7nHYQiOEC
Main work by Kike Portales, kudos to all coauthors!
Work supported by @FDL_Europe@trillium_tech and @isp_uv_es
Two months ago we published a follow-up of WorldFloods🌊 paper📜 https://t.co/x3knZjOOBc:
🗺️We curated and expanded the dataset to cover 144 floods
🌧️We trained a model to differentiate clouds and flood water
💻We tested the model on two large-scale floods in Pakistan & Australia
🧵 (1/4) "Semantic Segmentation of Methane Plumes with Hyperspectral Machine Learning Models" 🌈🛰️, was just published at the Nature Scientific Reports journal.
📜 Paper: https://t.co/Knhuym4Gc4
💻 Code: https://t.co/wFdjxrq5RZ
💾 Data: https://t.co/cNPgyZ46MZ
🆕 Cleaned, facelifted, updated our group webpage:
https://t.co/GA0XpNmSqD
✅List of most recent publications updated
✅ Last year's good news
✅ List of invited talks in 2023
✅ We're now 50 people! Welcome!
✅ A simple promo video w/ our activities
Feedback welcome! 😀
And Emiliano Díaz, now in his PhD final stage, today presents 'Learning causal drivers of PyroCb', showcasing how to make the most of heterogeneity to unveil causal triggers in phenomenons like photosynthesis and extreme wildfires! #EGU23@USMILE_ERC@FDL_Europe
Gonzalo Mateo García @gonzmg88 just defended his Ph.D. thesis "Transfer learning of deep learning models for cloud detection in optical satellite images" @isp_uv_es@UV_EG 🌍🛰️☁️🤖
The supervisor Luis Gomez-Chova, the group @isp_uv_es, family & friends are super-proud! 🥳😍
[Journal article] Proud to announce that our article "RaVÆn: unsupervised change detection of extreme events using ML on-board satellites" has just been published with @SpringerNature in @SciReports
Read here: https://t.co/VDfS79GRN7
#AI on-board @FDL_AI, @ESA_EO and @UniofOxford
😱 ¿Solo va 400.000 visualizaciones este vídeo? ¡Venga, vamos a quemar ese botón de retuit! 🔥🔥🔥 ¡Que llegue a la luna 🚀! ¡Y al despacho de la ministra @pilar_llop, que está más lejos!
📣 ¡¡¡LIBERAD LOS DATOS MERCANTILES!!!
Curious about the State of AI for EO?
Then read our new whitepaper/preprint w Cristian Rossi, Maral Bayaraa @SatAppsCatapult
An intro to
• sensors
• ML
• applications
To appeal to all backgrounds we included a handy glossary and acronym explainer.
https://t.co/JQSslESrcR
Taking uncertainty quantification in high-resolution traits estimation to the next level using #MachineLearning & @Google#earthengine. Check out Laura Martinez's new paper 📰 https://t.co/9xUinxhmZE and source code➡️ https://t.co/GHtbOjpRG9
The Universitat the València @UV_EG is the 11th worldwide (1st in Spain) in #remotesensing education according to Shanghai's index 2021
https://t.co/30HQUZUmtC
Congrats everyone teaching and learning!
Come and meet @remotekike He will be more than happy to explain his impressive advances in flood extent segmentation and cloud screening in Sentinel-2 images! @CopernicusEU@esa Is this a game changer in emergency response services? Must-see poster at #LPS22
Time flies but you can still apply for our PhD/postdoc positions.
Pasionate about #Machinelearning & #EarthSciences? Ready for great challenges?
Meet us in #LPS22 and #egu22 for a chat.