Introduction Several years ago, I completed my PhD research on the structure, chemistry, and electrodynamics of the ionosphere, with particular emphasis on the behaviour of the F₂‑region and the equatorial anomaly. Although this work was carried out in an earlier phase of space‑science research, the underlying physics, modelling approaches, and observational insights remain highly relevant today. With modern interest in satellite communication, GNSS (Global Navigational Satel

1. Introduction Counting fruits on trees is one of the most deceptively difficult problems in computer vision. Mangoes vary in size, shape, lighting, occlusion, and background clutter. Traditional image‑processing pipelines — thresholding, segmentation, contour detection — break down quickly in real orchards. Modern deep learning changed this landscape. Among all object‑detection architectures, YOLO (You Only Look Once) [1] stands out for its speed, simplicity, and real‑time

Revisiting My Early Ionospheric Modelling Work Through a Modern Lens Over the past several weeks, I have been restoring and digitizing some of my early research work from the mid‑1990s — work that originally existed only as printed pages filled with equations, derivations, and numerical modelling notes. Using a custom image‑processing pipeline that I developed recently (shadow removal, LAB‑space gradient correction, safe margin detection, and high‑accuracy OCR), these pages a

Coupled Oscillator Representation of Cardiac Electrical Activity The Synthetic ECG generation process begins with cardiac dynamics as network of biological oscillators. This network includes the Sinoatrial node (SA), the Atrioventricular (AV) node, and the His–Purkinje (HP) complex, each exhibiting an intrinsic rhythm and distinct electrical properties. Their ordered interaction gives rise to the heart's electrical signaling, guiding how these impulses propagate and enabling

Intuitus has released the Lepakshi-ECG Annotator Beta, a compact windows application designed for fast, accurate annotation of 12-lead ECG images. While the current version focuses on interactive bounding-box creation and JSON export, we are excited to preview a major breakthrough in our upcoming, yet-to-be-released version: the ability to use the software not just for annotation, but as high-fidelity grid removal engine for downstream Deep Learning models. The Lepakshi Workf

Introduction Ultrasound imaging is, at its core, an exercise in wave control . Every ultrasound image you have ever seen is the result of thousands of acoustic waves being launched, steered, focused, reflected, and coherently recombined in real time. The technique that makes this possible is beamforming . In this article, we take a structured journey through modern ultrasound beamforming: Analog vs digital beamforming FPGA‑ and GPU‑based architectures Why transmit and receive

Having spent years analyzing the medical ultrasound landscape from a patenting perspective , I’ve had a front-row seat to the intricate engineering that makes non-invasive imaging possible. However, the most complex systems often rely on the simplest physical truths. In this article, I’ve broken down the "black box" of ultrasound technology—starting with a simple "shouting in the hills" analogy and expanding into the sophisticated imaging modes that define modern diagnostics.

The term Software Defined Radio (SDR) was introduced by Joe Mitola in 1991 [1] to describe a new class of radios that are both programmable and reconfigurable. Unlike traditional radios, whose functionality is largely fixed by hardware components, an SDR uses software to implement many radio functions. As a result, the same physical hardware platform can support multiple communication standards, modulation schemes, and operating modes simply by changing the software, even du

The terrestrial ionosphere is one of the most fascinating natural plasma laboratories available to us. Stretching from roughly 60 km to more than 1000 km above Earth’s surface, it is sculpted by solar radiation, geomagnetic fields, and atmospheric chemistry. This region reflects and refracts radio waves, enabling long‑distance communication long before satellites existed. Its layered structure—D, E, and F regions—responds dynamically to solar activity, making it both a challe

A Brief Overview of Synthetic Data Synthetic data is information that is artificially generated by computer algorithms rather than collected from real-world events or individuals. While it contains no real personal records, it accurately mimics the mathematical patterns, distributions, and correlations of the original dataset. In the context of machine learning, this is typically achieved using generative AI techniques. For example, instead of relying on a database of highly

Intuitus has released Lepakshi- ECGAnnotator Beta version , a compact Windows application designed for fast, accurate annotation of 12‑lead ECG images. It supports interactive bounding‑box creation, keyboard‑driven adjustments, JSON export, and a clean, responsive UI — ideal for preparing datasets for YOLO (You Only Look Once) and other computer‑vision workflows. YOLO is a real‑time object detection method that identifies and localizes multiple objects in a single pass throug

Tavily Deep Research(*) is an AI-powered search engine designed specifically to act as the "eyes" for autonomous agents and large language models (LLMs). Unlike traditional search engines that prioritize visual layouts and advertisements for human browsing, Tavily focuses on parallel data extraction, scouring hundreds of web sources simultaneously to identify and aggregate the most relevant, high-quality information. By stripping away messy HTML "noise" and providing structur

In the current landscape of 2025, we are witnessing a fundamental shift in how the internet is used. For decades, the web was built for human eyes—optimized for visual layouts, advertisements, and sequential browsing. However, as AI agents become our primary researchers, a new infrastructure has emerged: Parallel Web Systems [1] Increasing Demand for Efficient Web Applications The need for quicker and more efficient web applications is increasing as users anticipate smooth ex

Searching is Science, but Finding is an Art This phrase is often repeated in the halls of Information Science, and for good reason. Searching is a mechanical act—the "Science. " It is the algorithm traversing billions of rows in a database, the cold logic of indexing, and the raw compute power that retrieves a list of results. Anyone can search; you simply type a keyword and hit "Enter." Finding , however, is the "Art." It is the human intuition that knows *how* to ask the q

If you’ve ever wondered how a research paper’s ideas ripple across disciplines, Litstudy[1] is the tool that brings those connections to life. Rather than simply tallying up citations, Litstudy dives deeper showing you exactly which domains have picked up on a paper’s concepts. To make the experience even more interactive, I’ve included a Python notebook in HTML format by taking the BiomedCLIP [2] paper as an example. BiomedClip Model is featured in AI Kosh, the Government of

In the current high-performance computing environment, developers require tools that are both powerful and easy to configure. Fortran and CUDA remain essential for scientific and engineering applications but integrating them can be complex and error prone. Docker addresses this issue by offering a consistent, portable environment that packages code, compilers, libraries, and system tools, thus resolving the typical “it works on my machine” problem. This blog post builds on r

** FORTRAN code work is in progress will also be shared from GitHub References Geomagnetic coordinates, time, and field in Centered and Eccentric Dipole approximations. Ramana, K V V , Murthy, K S R N and Khan, Ibrahim , Indian Journal of Radio & Space Physics Vol. 27, February 1998, pp.35-42. http://nopr.niscpr.res.in/handle/123456789/35273 A Theoretical Study of F2-region Equatorial Anomaly in Solar Maximum and Minimum, K S R Narayana Murthy, PhD (Thesis), 1995. Departmen

In this blog post, I discuss my preliminary experiences with a speaker-adaptive text-to-speech (TTS) system aimed at producing...

Introduction In my previous blog post, I discussed the input data for the LSTMVAE model, which comes from the publicly available PTB-XL...

Introduction In this post, I explore a custom LSTM-based Variational Autoencoder (LSTMVAE) designed for sequential representation...