Hi! This tutorial includes my research, presentations, slides, recommendations and resources for the remote sensing area. It will be updated and renewed over time. I suggest you to star and follow the developments in the repo.
Object detection is a technology related to computer vision and image processing that allows us to detect, identify, classify and track objects in images and videos. Many objects can be detected with object detection techniques. Plane detection, bridge detection, vehicle detection, ship detection and more can do with the object detection. In the image below, there are some examples of remote sensing data. (Image Source: https://www.mdpi.com/2072-4292/11/3/339)
- COCO (Common Objects in Context)
- Pascal VOC()
- ISPRS datasets: semantic labeling, reconstruction [https://www.isprs.org/data/]
- Toronto Massachusetts Roads and Buildings Dataset [https://www.cs.toronto.edu/~vmnih/data/]
- IEEE GRSS Data Fusion Contests: [http://www.grss-ieee.org/community/technical-committees/data-fusion/data-fusion-contest/]
- IEEE GRSS: hyperspectral datasets with standard train/test splits (DFC2018, Pavia, Indian Pines) [http://dase.grss-ieee.org/]
- INRIA Aerial Semantic labeling dataset: buildings [https://project.inria.fr/aerialimagelabeling/]
- XView: objects in aerial images [http://xviewdataset.org/]
- DOTA: Detecting Objects in Aerial images [https://captain-whu.github.io/DOTA/dataset.html]
I made a 4 week webinar on this subject. It was for the beginner level on deep learning. The slides and notes were in English, the webinar videos were in Turkish. All slides will be converted to cheat sheet and detailed notes will be added. You can find the presentations and slides below.
This episode made for beginners on Deep Learning and has suggestions, tricks and more.
Here, You can reach the webinar for Episode 1 that I gave with UHUZAM. The webinar language is Turkish.
You can find the presentation slide here. The slide language is English.
1-AI, ML, DL, Data terms and differences
2-Computer Vision ıntroduction
3-Computer Vision Usage Areas
4-Types of Deep Learning
5-Creating Model - Example CNN
6-DL Frameworks and Libraries
7-DL Working Environments
8-Suggested Resources
This episode made for the second week and beginners. After the first week, I wanted to talk about Tensorflow, other frameworks, examples for beginner and deep learning project
Here, You can reach the webinar for Episode 2 that I gave with UHUZAM. The webinar language is Turkish.
You can find the presentation slide here. The slide language is English.
1-What is Tensorflow?
2-Why we use Tensorflow?
3-What is Tensorflow differences for other frameworks?
4-Github Usage - Source Code search, find and review
This episode made for especially remote sensing and deep learning. After the first two week, I wanted to give about remote sensing with deep learning and make the segmentation example
Here, You can reach the webinar for Episode 3 that I gave with UHUZAM. The webinar language is Turkish.
You can find the presentation slide here. The slide language is English.
1-Some Examples in Remote Sensing and Deep Learning
2-Data Basics (Various Data Types (Hyperspectral - SAR etc))
3-ML and DL Based Object Detection
4-Object Detection Algorithms
5-Object Detection Dataset
6-Some Good Sources for Segmentation, Object Detection etc based process
7-Practical Session and Sources
Here, You can reach the webinar for Episode 4 that I gave with UHUZAM. The webinar language is Turkish.
You can find the presentation slide here. The slide language is English.
1-Capturing a 3D World
2-3D Problems and Deep Learning Techniques
3-3D Cloud Point
4-Deep Learning on 3D Point
5-Public Datasets for 3D