Harpreet Sahota 🥑’s Post

This is amazing technology that can be used in so many situations.

LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding Shen et al.: https://lnkd.in/eseBYPyk #Artificialintelligence #Deeplearning #MachineLearning

https://lnkd.in/d9bKAJRV The tech to create semantic objects and spatial relationships from basic scan data is coming really quickly now. What you do with that and how you can exploit that is one of the areas we're looking at. A volume of space that is tagged and has values, is very voxelly. Having played with Gaussians and Nerfies and photogrammetry. They are still inherently hollow models. Meaning within the volume lets you consider how to augment, improve or analyse.

Explore Object Detection from Satellite Images using Computer Vision with Roboflow! Eager to harness the power of computer vision for satellite imagery? Check out our latest tutorial: "Object Detection From Satellite Images using Computer Vision with Roboflow"! Watch Now: https://lnkd.in/ezDwP7He In this comprehensive guide, we'll take you through the steps to perform object detection on satellite images using the Roboflow platform. Learn how to detect various objects from space with accuracy and efficiency! What's Covered: Introduction to Satellite Image Object Detection Dataset Preparation and Annotation Leveraging Roboflow for Image Preprocessing Training Your Object Detection Model Evaluating and Testing Your Model Fine-tuning for Enhanced Accuracy Real-World Applications of Satellite Image Detection Ready to dive into satellite image object detection with Roboflow? Click the link above to start watching the tutorial now! Don't forget to like, subscribe, and share with your fellow enthusiasts for more cutting-edge tutorials. Share this post with your network and let's explore the possibilities of satellite image detection together! #SatelliteImages #ObjectDetection #ComputerVisionTutorial #Roboflow #pyresearch

Read how DUG Multi-parameter FWI Imaging breathes new life into vintage data, in this new feature on GEO EXPRO! https://lnkd.in/ge5cn2Jv

A new study has been published concerning the determination of aerosol properties using images from an all-sky camera. The presented approach works in all sky conditions, but with focus partially cloudy scenes. A Gaussian Process Regression (GPR) machine learning has been used to associate RGB radiances with aerosol optical depth (AOD) and Angstrom exponent (AE) observed by AERONET. The method has been tested on a 2-year dataset with varying atmospheric conditions. AOD and AE predictions showed excellent agreement with AERONET measurements, particularly when using the quality assurance method.. Sensitivity analysis confirms the stability of our methodology, especially when using quality assurance criteria. This approach could potentially enhance the use of all-sky cameras in studying aerosol-cloud interactions in partially cloudy scenarios. https://lnkd.in/dbRSTzz4

Pretrained self-supervised models have greatly advanced many applications in remote sensing. However, satellite images have some unique differences compared to your everyday photograph. The inclusion of multiple spectral bands and images from different time points tell a more complete story than one RGB image alone. Xuyang Li et al. developed Spatial-SpectralMAE (S2MAE) to model these unique aspects with a masked autoencoder. By using both spectral and spatial information, the model learns to reconstruct the input image when 90% of it is masked out. Experiments demonstrated that S2MAE outperforms other alternatives for land cover/use and change detection. Ablation studies further revealed the most critical components of S2MAE: 1) A high masking ratio 2) The use of 3D random masking for spectral data 3) Progressive pretraining on diverse datasets https://lnkd.in/eipqjTyD #RemoteSensing #EarthObservation #MachineLearning #DeepLearning #ComputerVision __________________ Enjoyed this post? Like 👍, comment 💬, or re-post 🔄 to share with others. Click "View my newsletter" under my name ⬆️ to join 1500+ readers.

#11 Seg-Eval Choosing the proper ground sampling distance is a pivotal decision in remote sensing downstream applications. In this work, the authors try to set out a clear ensemble of guidelines for fairly comparing semantic segmentation results obtained at various spatial resolutions. They also propose region-based pixel-wise metrics, allowing for a more detailed analysis of the model performance. what I liked: very practical and important problem paper: https://lnkd.in/gFfUE35v #50papers #AI4EO #remotesensing #deeplearning #GeoAI

Adaptive Slicing-Aided Hyper Inference for Small #ObjectDetection in High-Resolution Remote Sensing #Images ✍️by Hao Zhang et al. 🔗Access the paper: https://brnw.ch/21wNea9 #objectdetection #remotesensing

#10 DOFA Multimodality is one of the keys to remote sensing foundation models. Most of the methods, however, focus on just one of them. DOFA employs an innovative approach utilizing wavelength as a unifying parameter across various EO modalities to achieve a more cohesive multimodal representation. DOFA is trained using a masked image modeling strategy, and a distillation loss is included to further optimize its performance. what I liked the most: the wavelength approach paper: https://lnkd.in/dTBSEHk4 #50papers #AI4EO #remotesensing #deeplearning #GeoAI