anupamarajkumar Goto Github PK

I am Anupama and welcome to my GitHub page. I am from India and currently, I am an MSc. graduate from Technische Universitaet, Berlin and ELTE, Budapest. I specialise in Autonomous Systems.

My interests lie at the intersection of computer vision and machine learning. I am keen on topics that involve making sense of the scenes coming from different types of sensors. I have as much interest in traditional image processing and computer vision as in deep learning. I always marvel at how deep learning is capable of making sense from images but I also believe that, garbage in is garbage out and that's why I love pre-processing the images using traditional methods and algorithms.
I love reading technical papers and try to implement and reproduce them as much as possible. Find implementations from some such papers in my 3D computer vision and sensor fusion repositories.
I believe as a researcher, we should not be limited by our coding capabilities. Hence, I try to practise as much coding as I can in my free time. Although I like C and C++ more, I have been pushing myself out of my comfort zone to get more familiar with Python. General C++ and data structure algorithms and PyTorch tutorial and implementations can be found in my repositories

A very interesting research topic that resonates with me, I am working on performing semantic segmentation on high-resolution satellite images to detect waste landfills. I am working on the project in association with the Hungarian Academy of Sciences, Budapest, Hungary and it is being supervised by Dr. Tamas Sziranyi and Dr. Andras Majdik.
In the absence of a high resolution dataset catering to landfills, I have been working on creating and annotating my own dataset. Currently I am in the process of training various machine learning models and once benchmarked, I would be happy to make the dataset available for further use and development! The information on the dataset, the implementation and results along with report is placed at : https://github.com/AnupamaRajkumar/LandfillDetection_SemanticSegmentation
The details about the specification of the dataset and how to request the dataset can be found here : https://towardsdatascience.com/introducing-a-very-high-resolution-dataset-of-landfills-and-waste-dumps-196d46d533fc
ESA Interview : https://earth.esa.int/eogateway/news/meet-a-young-researcher-who-combines-remote-sensing-with-deep-learning-techniques

I am planning to extend this project to include more advanced and varied deep learning architectures and studying their performance.

I am employed as a Sensor Fusion Engineer with Continental Automotive in Budapest. I work on developing sensor fusion algorithms to detect static objects in the driving environment.

1. Robust line and plane fitting using RANSAC - traditional and optimised implementation (C++) : https://github.com/AnupamaRajkumar/3D-Computer-Vision/tree/master/RANSAC
2. Image rectification and planar panoramic mosaicing using homography (C++): https://github.com/AnupamaRajkumar/3D-Computer-Vision/tree/master/HomographyEstimation
3. Structure from Motion (SfM) (C++) : https://github.com/AnupamaRajkumar/3D-Computer-Vision/tree/master/StructureFromMotion
4. Point registration algorithms-ICP and TrICP (C++): https://github.com/AnupamaRajkumar/Sensor-Fusion/tree/master/PointRegistration
5. Stereo matching techniques using naive and dynamic programmig (C++) : https://github.com/AnupamaRajkumar/Sensor-Fusion/tree/master/StereoMatching
6. Unsupervised clustering of point clouds (C++): https://github.com/AnupamaRajkumar/Sensor-Fusion/tree/master/PointCloudClustering
7. Guided image upsampling algorithms (C++): https://github.com/AnupamaRajkumar/Sensor-Fusion/tree/master/GuidedImageUpsampling
8. Basic filtering methods with performance metrics (C++): https://github.com/AnupamaRajkumar/Sensor-Fusion/tree/master/Filtering
9. Bilateral filtering (C++): https://github.com/AnupamaRajkumar/Sensor-Fusion/tree/master/BilateralFiltering
10. Prewitt Edge Detector (C++) : https://github.com/AnupamaRajkumar/ImageAndVideoAnalysis/tree/master/PrewittEdgeDetector_NMS
11. Otsu Thresholding (C++) : https://github.com/AnupamaRajkumar/ImageAndVideoAnalysis/tree/master/OtsuThresholding
12. Hough Transform for circles (C++): https://github.com/AnupamaRajkumar/ImageAndVideoAnalysis/tree/master/HoughTransform_Circles
13. OpenCV KLT Tracker and Dense Optical Flow (C++): https://github.com/AnupamaRajkumar/ImageAndVideoAnalysis/tree/master/MotionTracking_DenseOpticalFlow
14. Autoencoder for extracting features from SAR : https://github.com/AnupamaRajkumar/ProcessChainForPolSARFeature/tree/master/Modern_Features
15. Automatic detection of Landfill using Deep Learning : https://github.com/AnupamaRajkumar/LandfillDetection_SemanticSegmentation

1. I enjoy writing and maintain a medium blog where I keep posting from time to time although I would like to be more regular : https://anupamarajkumar.medium.com/ ✒️
2. I love reading, although I would like to read more fiction 📚
3. I thrive on travelling - visiting places, trying out new cuisines and meeting new people 🌏 🏖️
4. Currently in the pandemic, I am enjoying Netflix 📺

Let's connect on LinkedIn: https://www.linkedin.com/in/anupama-rajkumar-910a222a/