Satellite image deep learning

In this episode, I caught up with Nils Lehmann to learn about Uncertainty Quantification for Neural Networks. The conversation begins with a discussion on Bayesian neural networks and their ability to quantify the uncertainty of their predictions. Unlike regular deterministic neural networks, Bayesian neural networks offer a more principled method for providing predictions with a measure of confidence. Nils then introduces the Pytorch Lightning UQ Box project on GitHub, a tool that enables experimentation with a variety of Uncertainty Quantification (UQ) techniques for neural networks. Model interpretability is a crucial topic, essential for providing transparency to end users of machine learning models. The video of this conversation is also available on YouTube here* Nils’s website* Lightning UQ box on Github* Further reading: A survey of uncertainty in deep neural networksBio: Nils Lehmann is a PhD Student at the Technical University of Munich (TUM), supervised by Jonathan Bamber and Xiaoxiang Zhu, working on uncertainty quantification for sea-level rise. More broadly his interests lie in Bayesian Deep Learning, uncertainty quantification and generative modelling for Earth Observational data. He is also passionate about open-source software contributions and a maintainer of the Torchgeo package. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.satellite-image-deep-learning.com

In this episode I caught up with Samuel Bancroft to learn about segmenting field boundaries using Segment Anything, aka SAM. SAM is a foundational model for vision released by Meta, which is capable of zero shot segmentation. However there are many open questions about how to make use of SAM with remote sensing imagery. In this conversation, Samuel describes how he used SAM to perform segmentation of field boundaries using Sentinel 2 imagery over the UK. His best results were obtained not by fine tuning SAM, but by carefully pre-processing a time series of images into HSV colour space, and using SAM without any modifications. This is a surprising result, and using this kind of approach significantly reduces the amount of work necessary to develop useful remote sensing applications utilising SAM. You can view the recording of this conversation on YouTube here- Samuel on LinkedIn - https://github.com/Spiruel/UKFields Bio: Sam Bancroft is a final year PhD student at the University of Leeds. He is assessing future food production using satellite data and machine learning. This involves exploring new self- and semi- supervised deep learning approaches that help in producing more reliable and scalable crop type maps for major crops worldwide. He is a keen supporter in democratising access to models and datasets in Earth Observation and machine learning. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.satellite-image-deep-learning.com

In this episode I caught up with Yotam Azriel to learn about interpretable deep learning. Deep learning models are often criticised for being "black box" due to their complex architectures and large number of parameters. Model interpretability is crucial as it enables stakeholders to make informed decisions based on insights into how predictions were made. I think this is an important topic and I learned a lot about the sophisticated techniques and engineering required to develop a platform for model interpretability. You can also view the video of this recording on YouTube.* tensorleap.ai* Yotam on LinkedinBio: Yotam is an expert in machine and deep learning, with ten years of experience in these fields. He has been involved in massive military and government development projects, as well as with startups. Yotam developed and led AI projects from research to production and he also acts as a professional consultant to companies developing AI. His expertise includes image and video recognition, NLP, algo-trading, and signal analysis. Yotam is an autodidact with strong leadership qualities and great communication skills. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.satellite-image-deep-learning.com

In this episode I caught up with Daniele Rege Cambrin, to learn about Earthquake detection with Sentinel-1 (SAR) images. Daniele has a key role in organising a new competition on this task, SMAC: Seismic Monitoring and Analysis Challenge. The topics covered include the logistics of organising this competition, and the lessons Daniele learned from organising a previous one. You can also view the recording of this discussion on YouTube.- Daniele on LinkedIn- Competition websiteBio: Daniele Rege Cambrin is currently pursuing his Ph.D. and his research interests lie in deep learning. He is particularly interested in finding efficient and scalable solutions in areas such as remote sensing, computer vision, and natural language processing. Additionally, he has a keen interest in game development, and worked on two machine-learning competitions related to change detection. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.satellite-image-deep-learning.com

In this episode Robin catches up with Inon Sharony to learn about the fascinating world of machine learning with SAR imagery. The unique attributes of SAR imagery, such as its intensity, phase, and polarisation, provide rich information for deep learning models to learn features from. The discussion covers the innovative applications ASTERRA is developing, and the nuances of machine learning with SAR imagery. This video of this episode is available on YouTube* https://asterra.io/* https://www.linkedin.com/in/inonsharony/Bio: Inon Sharony is the Head of AI at ASTERRA, where he is responsible for pushing boundaries in the field of deep learning for earth observation. Sharony brings a decade of experience leading development of cutting-edge AI technology that meets real-world business and product needs. His previous roles include Algorithm Group Manager at Rail Vision Ltd and R&D Group Lead & Head of Automotive Intelligence at L4B Software. He was PhD trained in Chemical Physics at Tel Aviv University and combines his extensive academic background in Physics and his hands-on experience with machine learning to develop strategic AI solutions for ASTERRA. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.satellite-image-deep-learning.com

In this episode, Robin catches up up with Alistair Francis and Mikolaj Czerkawski to learn about Major TOM, which is a significant new public dataset of Sentinel 2 imagery. Noteworthy for its immense size at 45 TB, Major TOM also introduces a set of standards for dataset filtering and integration with other datasets. Their aim in releasing this dataset is to foster a community-centred ecosystem of datasets, open to bias evaluation and adaptable to new domains and sensors. The potential of Major TOM to spur innovation in our field is truly exciting. Note you can also view the video of this recording on YouTube here. The video also includes a demonstration of accessing the dataset and a walkthrough of the associated Jupyter notebooks.* Dataset on HuggingFace* PaperAlistair Francis is a Research Fellow at the European Space Agency’s Φ-lab in Frascati, Italy. Having studied for his PhD at the Mullard Space Science Laboratory, UCL, his research is focused on image analysis problems in remote sensing, using a variety of supervised, self-supervised and unsupervised approaches to tackle problems such as cloud masking, crater detection and land use mapping. Through this work, he has been involved in the creation of several public datasets for both Earth Observation and planetary science. Mikolaj Czerkawski is a Research Fellow at the European Space Agency’s Φ-lab in Frascati, Italy. He received the B.Eng. degree in electronic and electrical engineering in 2019 from the University of Strathclyde in Glasgow, United Kingdom, and the Ph.D. degree in 2023 at the same university, specialising in applications of computer vision to Earth observation data. His research interests include image synthesis, generative models, and use cases involving restoration tasks of satellite imagery. Furthermore, he is a keen supporter and contributor to open-access and open-source models and datasets in the domain of AI and Earth observation. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.satellite-image-deep-learning.com

In this video Robin catches up with Konstantin Klemmer to discus SatClip, which is a new global & general purpose location encoder trained on Sentinel 2 imagery. The conversation covered the training of encoders such as CLIP, and discussed the implications for downstream applications. Note you can also view the video of this recording on YouTube here* Konstantin on LinkedIn* SatCLIPBio: Konstantin is a postdoctoral researcher at Microsoft Research New England. His research interests lie broadly within geospatial machine learning and bridging adjacent domains like remote sensing or spatial statistics. Konstantin has a PhD from the University of Warwick and NYU, a Master's from Imperial College London and an undergraduate degree from the University of Freiburg, Germany. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.satellite-image-deep-learning.com

James Gallagher, technical marketer at Roboflow, discusses the latest AI innovations in image annotation, including models like Segment Anything and GroundingDINO. They explore automated labeling with CLIP and the use of Auto Distill for image annotation in pipelines. The podcast also covers the changing process of creating datasets and highlights resources like Roboflow.com for AI-powered image annotation.

A large fraction of acquired satellite images contain 2D projections of Earth. However, for many downstream applications, 3D understanding is beneficial or necessary. In recent years, deep learning has enabled a number of solutions for learning 3D representations from 2D satellite images. This episode delivers an overview of some of the prominent works in this area. Mikolaj hosts 3 guests: Dawa Derksen, Roger Marí, and Yujiao Shi, providing a summary of each guest’s contributions on the topic as well as a panel discussion. Note you can also view the video of this recording on YouTube hereDawa Derksen - Origins of Shadow-NeRF Dawa pursued a post-doctoral research fellowship at the European Space Agency from 2020-2022, and is currently working at the Centre National d’Etudes Spatiales (French Space Agency) where he is involved in the field of 3D Implicit Representation Learning applied to Remote Sensing. * 🖥️ Shadow-NeRFRoger Marí - EO-NeRF Roger is a post-doc researcher from Barcelona specialised in 3D vision tasks. He is currently working at the Centre Borelli, ENS Paris-Saclay, in France, where his research topic is the application of neural rendering methods to satellite image collections. He is the author of Sat-NeRF and EO-NeRF, some of the first models in the literature to provide quantitatively convincing results in terms of surface reconstruction.* 🖥️ https://rogermm14.github.io/* 🖥️ EO-NeRFYujiao Shi - Connecting Satellite Image with StreetViewYujiao is a research fellow at the Australian National University. She obtained her PhD degree at the same institute. Her research interests include satellite image-based localisation, cross-view synthesis, 3D vision-related tasks, and self-supervised learning.* 🖥️ https://shiyujiao.github.io/* 📖 Geometry-Guided Street-View Panorama Synthesis from Satellite ImageryHost & Production: Mikolaj Czerkawskihttps://mikonvergence.github.io This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.satellite-image-deep-learning.com

In this episode Robin catches up with Jake Wilkins to learn about Deep learning in Google Earth Engine. Jake has been building commercial Earth Engine applications for the past three years and in this conversation he describes the pros and cons of several approaches to using deep learning models with Earth Engine. Note you can also view the video of this recording on YouTube here* Jake on LinkedIn* https://earthengine.google.com/Bio: Jake is a Software Engineer and Data Scientist based in London, UK. He has been building commercial Google Earth Engine applications for the past three years. His significant contributions include the no-code platform, Earth Blox, and the climate monitoring platform STRATA for UNEP (United Nations Environmental Programme). Alongside this, Jake has consistently developed his skills in machine learning, and a notable accomplishment in this field is winning the Earth-i hackathon last year. Jake has a deep passion for addressing the climate crisis and is committed to making Earth Observation more accessible to combat it. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.satellite-image-deep-learning.com