satellite-image-deep-learning / techniques Goto Github PK
View Code? Open in Web Editor NEWTechniques for deep learning with satellite & aerial imagery
License: Apache License 2.0
Techniques for deep learning with satellite & aerial imagery
License: Apache License 2.0
Many github links return Status: 429
Todo: ignore these
Great list, have suggested it to multiple people that are looking for DL+Geo ressources.
Solaris should def. be added here, great tool! DL framework for remote sensing data a bit similar to rastervision but with some unique very vool features! https://github.com/CosmiQ/solaris
On mobile for some time so no PR sorry :)
To catch bad links. Options
Dear Sir,
According to our last conversation, you were looking for a collapsible markdown format for the readme file. A format is mentioned in the github doc. By using the <details>
tag, it is possible to create dropdown contents. I have tried to change the font styles within the <summary>
tag, but did not work.
AWS sponsors access
ERROR: 13 dead links found!
421
[✖] https://spacenetchallenge.github.io/datasets/datasetHomePage.html → Status: 404
422
[✖] http://srtm.csi.cgiar.org/srtmdata/ → Status: 0
423
[✖] https://neat-EO.pink → Status: 0
424
[✖] https://onlinelibrary.wiley.com/doi/10.1002/9781118724194.ch11 → Status: 403
425
[✖] https://www.esa.int/Applications/Observing_the_Earth/Monitoring_European_air_traffic_with_Earth_observation → Status: 403
426
[✖] https://elib.dlr.de/128117/1/SEN12MS_Preprint.pdf → Status: 401
427
[✖] https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2019JB017519 → Status: 403
428
[✖] https://moderndata.plot.ly/weather-maps-in-python-with-mapbox-gl-xarray-and-netcdf4/ → Status: 500
429
[✖] https://platform.digitalglobe.com/gbdx/ → Status: 404
430
[✖] https://course-v3.fast.ai/index.html → Status: 0
431
[✖] https://github.com/microsoft/aerial_wildlife_detection/tree/multiProject → Status: 404
432
[✖] https://www.datacamp.com/courses/working-with-geospatial-data-in-python?tap_a=5644-dce66f&tap_s=411670-1f1ebc → Status: 403
433
[✖] https://www.linkedin.com/in/robmarkcole/ → Status: 999
I want to add "DC-GANs" technique in "18. Generative Adversarial Networks (GANs)", with GitHub repo. & Research paper reference. Please create a PR for the same. Thank you.
We have recently released an updated version of the SEN1-2 dataset, it contains the same regions, but we have added all the S2 multispectral bands, All S1 bands as well as corresponding MODIS land cover.
Paper: https://elib.dlr.de/128117/1/SEN12MS_Preprint.pdf
Data: https://mediatum.ub.tum.de/1474000
It is now the largest, manually curated dataset of S1 and S2 products, with corresponding labels for land use/land cover mapping, SAR-optical fusion, segmentation and classification tasks
The README file in this repo has some bad links - [404:NotFound]
The markup version of the readme that is displayed for the main page in this repo contains the following links:
Status code [404:NotFound] - Link: https://www.kaggle.com/robmarkcole/satellite-image-classification (Example notebook)
Status code [404:NotFound] - Link: http://scikit-image.org/docs/dev/auto_examples/transform/plot_register_translation.html?highlight=cross%20correlation (Phase correlation in scikit-image)
Status code [404:NotFound] - Link: https://developers.planet.com/docs/quickstart/getting-started/ (API key)
Status code [404:NotFound] - Link: https://github.com/galiboo/olympus
(The link in the readme’s raw markdown may appear in a different format to the link above)
Theses bad links were found by a tool I very recently created as part of an new experimental hobby project: https://github.com/MrCull/GitHub-Repo-ReadMe-Dead-Link-Finder
I (a human) verified that this link is broken and have manually logged this Issue (i.e. this Issue has not been created by a bot).
If this has been in any way helpful then please consider giving the above Repo a Star.
Hi robmarkcole,
I was trying to implement the tutorial on YOLOv5 Object Detection for RarePlanes : https://github.com/jeffaudi/rareplanes-yolov5
But I found a bug which make one part of the tutorial no longer useable. It's related with AWS and CosmiQ.
The problem appear in step 3 of the A. AMI/EC2 part, when trying to find the AMI CosmiQ_YOLO_Planes. Which it was impossible to find. Here you can have a look at my AWS AMI search.
After some research I found out that CosmiQ has been closed down since March 2021 : https://medium.com/the-downlinq/closing-time-cosmiq-works-is-closing-down-and-ending-its-leadership-of-spacenet-a53ba239745b
Therfore, all of the AMI they provided have been removed. So this part of the tutorial is not longer useable.
However, I could still download the RarePlanes dataset from the s3Bucket to work locally and accomplish the second part : B. On your own GPUs.
Thanks for your help with this wonderful repo,
Alban Tchikladzé
Currently datasets live there but that is not the main focus of this repo
Appears I have so many github links the Ci is raising Status: 429
codes and failing the run
Object Segmentation Foreground-Aware Relation Network for Geospatial Object Segmentation in High Spatial Resolution Remote Sensing Imagery (https://github.com/Z-Zheng/FarSeg) (CVPR 2020).
Hyperspectral Image Segmentation FPGA: Fast Patch-Free Global Learning Framework for Fully End-to-End Hyperspectral Image Classification (https://github.com/Z-Zheng/FreeNet) (TGRS 2020).
All-Weather Mapping Deep multisensor learning for missing-modality all-weather mapping (https://www.sciencedirect.com/science/article/pii/S0924271620303476) (ISRPS P&RS 2021).
Object Detection HyNet: Hyper-scale object detection network framework for multiple spatial resolution remote sensing imagery (https://www.sciencedirect.com/science/article/pii/S0924271620301167) (ISRPS P&RS 2020)
👋 Hi @robmarkcole thanks so much for compiling these resources. Just wanted to let you know that PEARL is publicly available for free. Could it maybe please be moved from the useful paid software
section into the semantic segmentation section
?
Hello! I want to convert the optical images of SEN1-2 to grayscale and then to dB scale. However, using just the dB formula (20*log10) gives much brighter images than the SAR ones.
Can you please describe how was the dB scale of SAR images calculated?
I'm looking for DSM satalllite images, is there any of them?
An additional dataset which is specifically made for deep learning on SAR and optical imagery is the SEN1-2 dataset. Contains around 250K corresponding patch pairs of Sentinel 1 (VV) and 2 (RGB) data.
Data download: https://mediatum.ub.tum.de/1436631
Paper: https://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/IV-1/141/2018/isprs-annals-IV-1-141-2018.pdf
It would be nice to hit a button and expand all links, so the page could be searched without needing to dive into the raw markdown
[✖] https://github.com/aws-samples/amazon-rekognition-video-analyzer → Status: 429
415
[✖] https://github.com/RemotePixel/remotepixel-api → Status: 429
416
[✖] https://github.com/zarr-developers/zarr → Status: 429
417
[✖] https://github.com/PacktPublishing/-Practical-Deep-Learning-on-the-Cloud → Status: 429
418
[✖] https://github.com/jrosebr1/simple-keras-rest-api → Status: 429
419
[✖] https://github.com/pytorch/serve → Status: 429
420
[✖] https://github.com/ryfeus/amazon-rekognition-custom-labels-satellite-imagery → Status: 429
421
[✖] https://github.com/developmentseed/chip-n-scale-queue-arranger → Status: 429
422
[✖] https://github.com/developmentseed/fastai-serving → Status: 429
423
[✖] https://github.com/Esri/raster-deep-learning → Status: 429
ICCV21, Change is Everywhere: Single-Temporal Supervised Object Change Detection in Remote Sensing Imagery
RSE21, Building Damage Assessment for Rapid Disaster Response with a Deep Object-based Semantic Change Detection Framework: from natural disasters to man-made disasters
ISPRS P&RS 22, ChangeMask: Deep Multi-task Encoder-Transformer-Decoder Architecture for Semantic Change Detection
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.