Prof for AI and Data Science for Earth Observation
@TU_muenchen
, PI of
@zhu_xlab
and
@ai4eo_de
, co-director Munich Data Science Institute - opinions are my own
Interested in land use dynamics on the Earth's surface? Check our
#DynamicEarthNet
dataset -- daily, multi-spectral satellite observations of 75 AOIs across the globe with imagery from
#Planetscope
and
#Sentinels
, paired with high quality
#LULC
labels
#AI4EO
#CVPR
How to make your model better transferrable across geographical regions?
Take a look at our recent paper
@Zhu_XLab
@ai4eo_de
led by Qingyu who proposed CrossGeoNet for
#buildingfootprint
extraction
link to paper:
Interested in training your own vision language foundation models
#VLM
in
#EarthObservation
? We curated
#ChatEarthNet
- a
#global
&
#highquality
dataset with paired satellite image, land cover map and the corresponding text description.
Link to paper:
We would be happy to host
#Urkaine
scientists who is working on
#AI4EO
in our future lab
@ai4eo_de
as beyond fellows (=funded stay of 3~6months). Please kindly spread the word. For interested individuals, please don’t hesitate to contact us ai4eo
@tum
.de
@IEEE_GRSS
@HIDAdigital
How significant would
#urbanization
impact urban tree coverage? Jianhua extracted
#UrbanTreeCoverage
of more than 800 cities in North America via
#AI4EO
, from which prevalent positive indirect effect of urbanization on UTC was revealed.
Link to our paper:
#Multi
-scene datasets are too labor-consuming to generate?
check our recent work
@Zhu_XLab
: we recognize multiple scenes in a single image by leveraging massive well-annotated single-scene images.
#AI4EO
Link to paper:
Interested in a new
#AI4EO
task - multi-scene classification in single aerial images?
We
@Zhu_XLab
@ai4eo_de
are sharing a large-scale benchmark, called
#MultiScene
, composed of 100,000 high-resolution aerial images. Stay tuned!
Link to paper:
🚀 Perspectives on Earth and Climate Foundation Models! 🌍 Check out our new paper, "On the Foundations of Earth and Climate
#FoundationModels
," where we define 11 essential features to guide research towards the ideal FM.
👉 Read it here:
Can
#DeepLearning
based
#CloudRemoval
methods
#generalize
to a diverse set of observations acquired at variable cloud coverage or in different regions and seasons?
Try our model or test our open data
#SEN12MS
-CR.
Paper:
Data:
To what extent does self-supervised learning help in learning generic feature representation from the big EO data achieve and support downstream tasks in a label-efficient way?
Take a look at our survey article
@zhu_xlab
led by Yi:
#SSL
#AI4EO
@ai4eo_de
Semantic segmentaion with sparse scribbled annotation?
Check our recent work
@zhu_xlab
, in collaboration with Diego Marcos and
@devistuia
paper:
Data & code:
How to generate 3D city maps from single aerial/satellite images? How can we profit from synthetic dataset and transfer the learned representation to real world scenarios?
Check our recent work
@Zhu_XLab
@ai4eo_de
:
paper:
codes:
Congratulations to
@LloydHughesZA
, who has successfully defended his PhD today on the topic of
#DeepLearning
for
#SAR
and
#optical
image matching. Many congrats to Michael as well, as the doctor father. Many thanks to Martin and Florence for being part of it.
#AI4EO
@Zhu_XLab
Am 1.5. beginnen 3 internationale Zukunftslabore
#K
ünstlicheIntelligenz in
#Berlin
,
#Hannover
&
#M
ünchen ihre Forschung zu aktuellen Fragen der
#KI
. Sie sind aus einem
#BMBF
Wettbewerb hervorgegangen & mit jeweils bis zu 5 Mio. € gefördert. Zur PM 👉
#KI
Looking for a motivated individual to team up with me as my deputy at TUM-SiPEO.
What we offer: to co-organize my lab consisting of a young and motived team on
#AI4EO
for
#socialgood
, while still being able to carry out independent research.
pls RT
Tweeting for
#socialgood
? We discuss how we could harvest geo-information from
#socialmedia
data, such as text messages and images: perspectives, data issues, geo-information retrieval,
#ethical
opportunities and applications.
#remotesensing
Link:
Looking for a
#PhD
program to train you to become an outstanding Data Scientist? 🎓 💻
👉 Join the Munich School for
#DataScience
and enjoy innovative supervision concepts and profit from a thriving environment for young scientists 🤓
💥
The results of the latest ERC
#ProofOfConcept
grant competition are out.
55 researchers received €150,000 each to explore the commercial or societal potential of the results of their research projects.
Curious? 👉
#ERCPoC
#FrontierResearch
This week's highlight of
@Zhu_XLab
: Very proud to see 8 articles from my lab are among 50 most popular papers
@IEEE_GRSS
#TGRS
:
and two among the most popular five:
Interested in getting an overview on
#building
extraction from
#RemoteSensing
imagery? Qingyu complied a comprehensive survey, recently published
@IEEE_GRSS
#TGRS
- RS offers a perspective to complement incomplete
#OSM
building data.
Link to paper:
Did you know the urban tree canopy coverage in Brazil is between 5% and 35%? Jianhua made nationwide mapping and coverage assessment in Brazil from high-resolution
#remotesensing
images using
#deeplearning
.
#AI4EO
Link to our paper
@isprs
:
How often do you receive rejections based on reviewers who haven’t really understood what you are doing? Seem to be unavoidable experience if long enough in academia…
Exciting news: Today Lukas
@kondmann
has successfully defended his PhD on the topic of
#RemoteSensing
#TimeSeries
data analysis! It was a great time having him
@Zhu_XLab
. Many thanks to Profs U. Hugentobler, M. Schmitt and D. Tuia
@devistuia
for serving in the committee.
The Munich Center for Machine Learning is calling for PhD applications:
Interested in working with us
@Zhu_XLab
@ai4eo_de
on
#AI4EO
? Please kindly send in your application and choose us as the PIs you want to work with.
Congrats to our brand new Dr. Yuansheng Hua
@Zhu_XLab
who successfully defended his PhD thesis today on
#DeepLearning
for Aerial Scene Understanding from the Lab to the
#Wild
.
Follow him here:
Mega thanks to Sébastien Lefèvre who came to Munich for it.