From 9add597190594ca76e0687ad9208951a85cee1d9 Mon Sep 17 00:00:00 2001 From: Robin Date: Mon, 12 Sep 2022 06:27:51 +0100 Subject: [PATCH] Update README.md --- README.md | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/README.md b/README.md index f509ed06..0fdf3155 100644 --- a/README.md +++ b/README.md @@ -193,6 +193,7 @@ or [fastai](https://medium.com/spatial-data-science/deep-learning-for-geospatial * [droughtwatch](https://github.com/wandb/droughtwatch) -> code for 2020 [paper](https://arxiv.org/abs/2004.04081): Satellite-based Prediction of Forage Conditions for Livestock in Northern Kenya * [JSTARS_2020_DPN-HRA](https://github.com/B-Xi/JSTARS_2020_DPN-HRA) -> code for 2020 [paper](https://ieeexplore.ieee.org/document/9126161): Deep Prototypical Networks With Hybrid Residual Attention for Hyperspectral Image Classification * [SIGNA](https://github.com/kyle-one/SIGNA) -> code for 2022 [paper](https://arxiv.org/abs/2208.02613): Semantic Interleaving Global Channel Attention for Multilabel Remote Sensing Image Classification +* [Satellite Image Classification](https://github.com/rocketmlhq/rmldnn/tree/main/tutorials/satellite_image_classification) using rmldnn and Sentinel 2 data ## Segmentation Segmentation will assign a class label to each **pixel** in an image. Segmentation is typically grouped into semantic, instance or panoptic segmentation. In semantic segmentation objects of the same class are assigned the same label, whilst in instance segmentation each object is assigned a unique label. Panoptic segmentation combines instance and semantic predictions. Read this [beginner’s guide to segmentation](https://medium.com/gsi-technology/a-beginners-guide-to-segmentation-in-satellite-images-9c00d2028d52). Single class models are often trained for road or building segmentation, with multi class for land use/crop type classification. Image annotation can take longer than for object detection since every pixel must be annotated. **Note** that many articles which refer to 'hyperspectral land classification' are actually describing semantic segmentation. Note that cloud detection can be addressed with semantic segmentation and has its own section [Cloud detection & removal](https://github.com/robmarkcole/satellite-image-deep-learning#cloud-detection--removal) @@ -253,6 +254,7 @@ or [fastai](https://medium.com/spatial-data-science/deep-learning-for-geospatial * [A2-FPN](https://github.com/lironui/A2-FPN) -> code for 2021 [paper](https://arxiv.org/abs/2102.07997): A2-FPN for Semantic Segmentation of Fine-Resolution Remotely Sensed Images * [MAResU-Net](https://github.com/lironui/MAResU-Net) -> code for 2020 [paper](https://arxiv.org/abs/2011.14302): Multi-stage Attention ResU-Net for Semantic Segmentation of Fine-Resolution Remote Sensing Images * [ml_segmentation](https://github.com/dgriffiths3/ml_segmentation) -> semantic segmentation of buildings using Random Forest, Support Vector Machine (SVM) & Gradient Boosting Classifier (GBC) +* [RSEN](https://github.com/YonghaoXu/RSEN) -> code for 2021 [paper](https://arxiv.org/abs/2104.03765): Robust Self-Ensembling Network for Hyperspectral Image Classification ### Segmentation - Land use & land cover * [nga-deep-learning](https://github.com/jordancaraballo/nga-deep-learning) -> performs semantic segmentation on high resultion GeoTIF data using a modified U-Net & Keras, published by NASA researchers @@ -971,6 +973,7 @@ Generally speaking, change detection methods are applied to a pair of images to * [FHD](https://github.com/ZSVOS/FHD) -> code for 2022 [paper](https://ieeexplore.ieee.org/document/9837915): Feature Hierarchical Differentiation for Remote Sensing Image Change Detection * [Change detection with Raster Vision](https://www.azavea.com/blog/2022/04/18/change-detection-with-raster-vision/) -> blog post with Colab notebook * [building-expansion](https://github.com/reglab/building-expansion) -> code for 2021 [paper](https://arxiv.org/abs/2105.14159): Enhancing Environmental Enforcement with Near Real-Time Monitoring: Likelihood-Based Detection of Structural Expansion of Intensive Livestock Farms +* [SaDL_CD](https://github.com/justchenhao/SaDL_CD) -> code for 2022 [paper](https://arxiv.org/abs/2205.13769): Semantic-aware Dense Representation Learning for Remote Sensing Image Change Detection ## Time series More general than change detection, time series observations can be used for applications including improving the accuracy of crop classification, or predicting future patterns & events. Crop yield is very typically application and has its own section below @@ -1465,6 +1468,7 @@ Data fusion covers techniques which integrate multiple datasources, for example * [HSHT-Satellite-Imagery-Synthesis](https://github.com/yuvalofek/HSHT-Satellite-Imagery-Synthesis) -> code for thesis - Improving Flood Maps by Increasing the Temporal Resolution of Satellites Using Hybrid Sensor Fusion * [MDC](https://github.com/Kasra2020/MDC) -> code for 2021 [paper](https://ieeexplore.ieee.org/document/9638348): Unsupervised Data Fusion With Deeper Perspective: A Novel Multisensor Deep Clustering Algorithm * [FusAtNet](https://github.com/ShivamP1993/FusAtNet) -> code for 2020 [paper](https://ieeexplore.ieee.org/document/9150738): FusAtNet: Dual Attention based SpectroSpatial Multimodal Fusion Network for Hyperspectral and LiDAR Classification +* [AMM-FuseNet](https://github.com/oktaykarakus/ReSIF/tree/main/AMM-FuseNet) -> code for 2022 [paper](https://www.mdpi.com/2072-4292/14/18/4458): AMM-FuseNet: Attention-Based Multi-Modal Image Fusion Network for Land Cover Mapping ## Terrain mapping, Disparity Estimation, Lidar, DEMs & NeRF Measure surface contours & locate 3D points in space from 2D images. NeRF stands for Neural Radiance Fields and is the term used in deep learning communities to describe a model that generates views of complex 3D scenes based on a partial set of 2D images @@ -2689,6 +2693,8 @@ Many datasets on kaggle & elsewhere have been created by screen-clipping Google * [landsatlinks](https://github.com/ernstste/landsatlinks) -> A simple CLI interface to generate download urls for Landsat Collection 2 Level 1 product bundles * [pyeo](https://github.com/clcr/pyeo) -> a set of portable, extensible and modular Python scripts for machine learning in earth observation and GIS, including downloading, preprocessing, creation of base layers, classification and validation. * [metaearth](https://github.com/bair-climate-initiative/metaearth) -> Download and access remote sensing data from any platform +* [geoget](https://github.com/mnpinto/geoget) -> Download geodata for anywhere in Earth via ladsweb.modaps.eosdis.nasa.gov +* [geeml](https://github.com/Geethen/geeml) -> A python package to extract Google Earth Engine data for machine learning ## Image augmentation packages Image augmentation is a technique used to expand a training dataset in order to improve ability of the model to generalise @@ -2750,6 +2756,7 @@ Image augmentation is a technique used to expand a training dataset in order to * [ODEON landcover](https://github.com/IGNF/odeon-landcover) -> a set of command-line tools performing semantic segmentation on remote sensing images (aerial and/or satellite) with as many layers as you wish * [aitlas-arena](https://github.com/biasvariancelabs/aitlas-arena) -> An open-source benchmark framework for evaluating state-of-the-art deep learning approaches for image classification in Earth Observation (EO) * [PaddleRS](https://github.com/PaddlePaddle/PaddleRS) -> remote sensing image processing development kit +* [RocketML Deep Neural Networks](https://github.com/rocketmlhq/rmldnn) -> read [Satellite Image Classification](https://github.com/rocketmlhq/rmldnn/tree/main/tutorials/satellite_image_classification) using rmldnn and Sentinel 2 data ## Model tracking, versioning, specification & compilation * [dvc](https://dvc.org/) -> a git extension to keep track of changes in data, source code, and ML models together