[{"content":"Publications This page is automatically updated from my ORCID profile.\nArticles Herman le Roux, Günther Drevin, Ruhann Steyn, Du Toit Strauss (2022), DEEP LEARNING APPROACH TO CLASSIFYING SOLAR RADIO BURSTS, Proceedings of the International Conferences on Applied Computing 2022 and WWW/Internet 2022. DOI Herman le Roux, Ruhann Steyn, Du Toit Strauss, Mark Daly, Peter T. Gallagher, Jeremiah Scully, Shane A. Maloney, Christian Monstein, \u0026amp; Günther Drevin (2025), Type II and Type III Solar Radio Burst Classification Using Transfer Learning, Solar Physics. DOI Peer Review Peer review for Journal of Space Weather and Space Climate - 2026 Presentations Manually maintained\nPosters \u0026ldquo;Segmentation of Solar Radio Bursts using Detectron2\u0026rdquo; presented at ML-Helio IHDEA - 2025 Last updated: 2026-04-06 08:23:18 UTC\n","permalink":"https://hermanlrx.github.io/Publications/","summary":"\u003ch1 id=\"publications\"\u003ePublications\u003c/h1\u003e\n\u003cp\u003eThis page is automatically updated from my \u003ca href=\"https://orcid.org/0000-0002-1805-0706\"\u003eORCID profile\u003c/a\u003e.\u003c/p\u003e\n\u003ch2 id=\"articles\"\u003eArticles\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003eHerman le Roux, Günther Drevin, Ruhann Steyn, Du Toit Strauss (2022),\n\u003cem\u003eDEEP LEARNING APPROACH TO CLASSIFYING SOLAR RADIO BURSTS\u003c/em\u003e,\n\u003cem\u003eProceedings of the International Conferences on Applied Computing 2022 and WWW/Internet 2022\u003c/em\u003e. \u003ca href=\"https://doi.org/10.33965/ac_icwi2022_202208l002\"\u003eDOI\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003eHerman le Roux, Ruhann Steyn, Du Toit Strauss, Mark Daly, Peter T. Gallagher, Jeremiah Scully, Shane A. Maloney, Christian Monstein, \u0026amp; Günther Drevin (2025),\n\u003cem\u003eType II and Type III Solar Radio Burst Classification Using Transfer Learning\u003c/em\u003e,\n\u003cem\u003eSolar Physics\u003c/em\u003e. \u003ca href=\"https://doi.org/10.1007/s11207-025-02595-w\"\u003eDOI\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2 id=\"peer-review\"\u003ePeer Review\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003ePeer review for Journal of Space Weather and Space Climate - 2026\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2 id=\"presentations\"\u003ePresentations\u003c/h2\u003e\n\u003cp\u003e\u003cem\u003eManually maintained\u003c/em\u003e\u003c/p\u003e","title":"Publications"},{"content":"Welcome to My Blog! It finally happened. I have now been made a cog in the machine that is GSoC.\nThis page will now exists as a platform for my peers and future colleagues to enjoy my GSoC journey. Whether it be failed pull requests, frustrating unsquashed commits or just simple grammar mistakes, I hope it will be helpful for all who endeavour to contribute to open source and academia.\nWhat is GSoC? GSoC (or Google Summer of Code) is a global online project run by Google. The focus of GSoC is to attract contributors to the open-source software space. By allowing organisations to sumbit projects and accompanying mentors, GSoC empowers newcomers to the contribute to the vast and unknown world of open source software projects.\nProject Selection The project that I have selected to apply for is called radiospectra. Radiospectra is a Python software package for reading and plotting radio spectrogram data from various ground and space based solar radio observatories. Radiospectra also forms a part of the SunPy library, a open-source solar data analysis environment.\nDuring my 4th year research project I was suddenly faced with data that I had never worked with before! What is a .fit file? Why do they have headers? Why can\u0026rsquo;t I just use plt.plot(fitFile)? All wonderful questions that I did not have answers to. But the human spirit prevailed, or at least in that case radiospectra did.\nUsing radiospectra I was able to plot e-CALLISTO data that would be used as part of my Masters project later on. At the time it was, however, clear that radiospectra lacked some useful features. But to my mind all of these libraries were all created and packaged in these far away lands by coding fairies that sleep very little and see the Sun even less. So in the end I went my own way coding things as I needed them without real regard for future use.\nJumping to the present, I now find myself in a position where I can contribute to a package that I have used since I started my academic career! The most shocking part of it all? I could have been contributing that entire time.\nThe big lesson here is that it really is just that easy to start working on something! You just have to go looking in the right places and read a whole lot of docs.\nAs part of my PhD I will be working with solar radio data from the I-LOFAR radio station and hopefully other data sources in the future as well. Contributing and keeping track of the radiospectra package has become more important than ever as it will streamline a lot of my workflow. Hopefully community members using radiospectra will benefit long term as well.\nWhat You Can Expect I will be keeping blogs focused on my recent GSoc 2026 application process. Should I be accepted as a GSoC contributor for radiospectra, then there will be weekly updates kept on this blog.\nWhile the goal of keeping blogs to have a publically accessible record of project goals, accomplishments and remaining tasks, I hope to offer a more personal view on the highs and lows of the journey. As the contributor guide aptly states: \u0026ldquo;If you think you know everything, then you are not good enough for GSoC!\u0026rdquo;.\nFrom my point of view this can not be more true! It is simply the case that few university programs in the world a can tailor a program for students in such a way that they are entirely prepared for anything a new code base throws at them. To me, that makes this project all the more exciting. No experience is as valuable as hands on development and getting to interact with contributors who have been a part of the open source community.\nSo to ensure that readers get a representative view of what the GSoC experience is like, especially for someone who has a prior knowledge and experiences in coding, the blog posts will not only include project updates but also some of the hilarious pitfalls, mistakes and general misery that a lack of thorough reading can cause.\nWhat\u0026rsquo;s Next? In my coming post I\u0026rsquo;m planning to write about my first pull request made to SunPy/radiospectra.\nCoden snapshot from my first PR! 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 elif (\u0026#34;SOLO\u0026#34; in cdf_globals.get(\u0026#34;Project\u0026#34;, \u0026#34;\u0026#34;)[0]): data_type = cdf_globals.get(\u0026#34;Data_type\u0026#34;, [\u0026#34;\u0026#34;])[0] data_descriptor = cdf_globals.get(\u0026#34;Descriptor\u0026#34;, \u0026#34;\u0026#34;)[0] if (\u0026#34;RPW-HFR-SURV\u0026#34; not in data_descriptor and \u0026#34;RPW-TNR-SURV-FLUX\u0026#34; not in data_descriptor): raise ValueError( f\u0026#34;Currently radiospectra supports Level 2 HFR survey data \u0026#34; \u0026#34;and Level 3 HFR, TNR survey data the file \u0026#34; f\u0026#39;{file.name} is {cdf_globals.get(\u0026#34;Logical_source_description\u0026#34;, [\u0026#34;\u0026#34;])[0]}\u0026#39; ) if(\u0026#34;L3\u0026#34; in data_type): epoch = cdf.varget(\u0026#34;Epoch\u0026#34;) times = Time(\u0026#34;J2000.0\u0026#34;) + epoch * u.ns freqs = cdf.varget(\u0026#34;FREQUENCY\u0026#34;) \u0026lt;\u0026lt; u.Unit( cdf.varattsget(\u0026#34;FREQUENCY\u0026#34;)[\u0026#34;UNITS\u0026#34;] ) My initial thought was that this would be a very simple job! I mean, how difficult can an elif be right? Keep a lookout for the next post to find out more!\nThanks for reading!\n","permalink":"https://hermanlrx.github.io/posts/gsoc2026_blogintro/","summary":"\u003ch2 id=\"welcome-to-my-blog\"\u003eWelcome to My Blog!\u003c/h2\u003e\n\u003cp\u003eIt finally happened. I have now been made a cog in the machine that is GSoC.\u003c/p\u003e\n\u003cp\u003eThis page will now exists as a platform for my peers and future colleagues to enjoy my GSoC journey. Whether it be failed pull requests, frustrating unsquashed commits or just simple grammar mistakes, I hope it will be helpful for all who endeavour to contribute to open source and academia.\u003c/p\u003e","title":"My First Blog Post"},{"content":"Hi, I\u0026rsquo;m Herman le Roux I am a PhD student at DIAS currently working on automated solar radio classification software using I-LOFAR data.\nWhy This Blog? I created this blog to share my GSoC experience.\nConnect With Me GitHub: @Hermanlrx This site is built with Hugo and the PaperMod theme.\n","permalink":"https://hermanlrx.github.io/about/","summary":"Learn more about me","title":"About"},{"content":"Herman le Roux Email: herman.leroux@dias.ie\nEducation Ph.D. in Computer Engineering | 2025 - Present Dublin Institute for Advanced Studies \u0026amp; Technological University of the Shannon\nDublin, Ireland\nResearch focus: Machine learning and image processing techniques for detection and classification of solar features and activity Chief Observer for Irish-LOFAR radio telescope (data management and observation scheduling) M.Sc. in Computer Science | 2023 - 2024 North-West University, Potchefstroom, South Africa\nDissertation: \u0026ldquo;Comparison of deep learning methods for solar radio burst classification\u0026rdquo; B.Sc. Honours in Computer Science and Information Systems | 2022 North-West University, Potchefstroom, South Africa\nGraduated with Distinction Specialisations AI, Image Processing, Decision Supports Systems and Databases. B.Sc. in Information Technology | 2019 - 2021 North-West University, Potchefstroom, South Africa\nGraduated with Distinction Coursera Specialization in Python 3 Programming (University of Michigan) Research Experience Ph.D. Researcher | 2025 - Present Dublin Institute for Advanced Studies\nDeveloping machine learning models for automated detection and classification of solar radio phenomena Managing data pipelines and observation scheduling for I-LOFAR radio telescope Conducting research in solar physics, radio astronomy, and deep learning applications Assisting masters research students in collecting and processing of data for M dwarf observations using I-LOFAR Research Supervision | 2023 - 2024 North-West University\nCo-supervised Honours research projects titled:\nGenerating type II and type III solar radio bursts using Generative Adversarial Networks Advancing Sunspot Measurement and Analysis: Integrating Deep Learning and Digital Image Processing Automated Calculation of Sunspot Area using Digital Image Processing Teaching Experience Temporary Lecturer | 2023 - 2025 North-West University\nArtificial Intelligence (CMPG 313) - 3rd year students (2023, 2024, 2025 distance cohort) Curriculum development, lecture delivery, assignment and exam creation and assessment Decision Support Systems (ITRI 628) - Graduate level (2024) Communications Skills (CMPG 214) - 2nd year IT/CS students (2023, 2024) Introduction to Programming with GUIs (CMPG 122) - 1st year students (2023) C# and Visual Studio Tutor | 2021-2022 North-West University\nObject-Oriented Programming (CMPG 211) - 2nd year students Introduction to GUIs (CMPG 122) - 1st year students Professional Development Awards \u0026amp; Honors Second-Best Paper Award - International Applied Computing Conference, Lisbon (2022) Best Research Project - B.Sc. Honours in IT \u0026amp; Computer Science cohort, North-West University (2022) Best Research Project Presentation - B.Sc. Honours in IT \u0026amp; Computer Science cohort, North-West University (2022) Outreach \u0026amp; Community Engagement DIAS Dunsink Observatory - Public outreach activities (2025-Present) Robotics Education - Community outreach for underprivileged children (2022-2024) ","permalink":"https://hermanlrx.github.io/CV/","summary":"My CV","title":"CV"}]