Ever wondered how you can control a game without ever touching a keyboard or controller? 🎮✋ Check out this quick sneak peek from our upcoming session! We’re excited to bring you another installment of Tea Tech Talks, and this time we’re diving deep into the fascinating world of Computer Vision! 🤖👁️ Join us for a fun, hands-on coding session where we’ll explore exactly how machines interpret images and movements. We’ll cover the basics of image processing and object detection through simple, interactive activities—just like the balloon blaster game in the video! 🎈💥 If you’re curious about AI or want to learn how computers “see” the world, this session is the perfect place to start. #ComputerVision #ArtificialIntelligence #MachineLearning #TechTalks #OpenCV #StudentDevelopers #TechCommunity #CodingEvent

🚀 Introducing EduSense — Smart Attendance with Face Recognition In today’s fast-paced digital world, traditional attendance methods feel outdated, inefficient, and prone to errors. That’s why I built EduSense, an intelligent attendance system that uses face recognition to automate and secure the entire process. 💡 What is EduSense? EduSense is a smart solution that detects and recognizes student faces in real time, marking attendance automatically—eliminating manual effort and preventing proxy attendance. 🔍 How it works: 📷 Captures live video through a camera 🧠 Detects and encodes facial features 🔐 Matches faces with a stored database ✅ Marks attendance instantly with timestamps ⚙️ Tech Stack: Python | OpenCV | Face Recognition | NumPy ✨ Why EduSense? ✔️ Saves time for teachers ✔️ Improves accuracy and reliability ✔️ Prevents fake attendance ✔️ Moves classrooms toward digital transformation This project reflects my interest in AI, computer vision, and building real-world solutions that make a difference. 🔗 GitHub Repository: https://lnkd.in/dgSghKeQ 📸 Project Demo: #AI #MachineLearning #ComputerVision #Python #OpenCV #StudentProjects #Innovation #EdTech #entracloud #superioruniversity

🎻 Strings Attached: A Rhythm-Based Game for Learning the Violin Bow 🎻 ❗ Latest Update: Rhythm-Synced Mechanics: Integrated real-time synchronization between physical bowing and character movement. ·      Purpose: To leverage natural pattern recognition, making practice feel more intuitive and engaging for students. ⚠️ Current Development Challenge: Technical limitations in OpenCV (the motion-tracking software). ·      Detail: The system currently struggles to distinguish between correct bowing and simple hand movement. ·      Goal: Refining the detection logic to ensure rewards are based on proper technique rather than general motion. 🔎 Future Roadmap: ·      Adaptive Difficulty: Implementing machine learning to adjust game difficulty based on how many successful "jumps" or "attacks" the user performs. ·      Inclusive Design: Adding toggles to disable flashing lights or metronome sounds for users with sensory sensitivities. 📒 Notes: ·   The song was made by me. You can hire me as your DJ for happy moments (ie. getting your manuscript accepted to a journal). ·   Also, the cool shades will not be included in the upcoming demo release sorry. #MusicTech #ComputerVision #EdTech #InteractiveTechnology #OpenCV #MediaPipe #Gamification

With AI we can’t hide behind knowing language syntax or knowledge of some specific API’s or frameworks anymore -although still important-. Ironically, that means fundamentals are more important than ever. Fundamental computer science stuff, like databases, system design, API design, operating systems, networking, containers and orchestrations, message queues, graphics rendering, game engines, even a little math like linear algebra and calculus….etc You don’t need to learn all of it obviously, you would need more than a lifetime anyway 😅. But diving deeper than surface level is not optional anymore. Stay hungry!

Just deployed V1 of Project 'Aegis-Reflect' at Sayyam AI Lab. I wanted to push the absolute limits of local, real-time Computer Vision and asynchronous execution using Python. 🚀 By leveraging Google Gemini as my engineering co-pilot, I was able to rapidly architect a highly-optimized, zero-latency AR deflection engine. 🔹 The Sensor Logic: Intercepting live webcam feeds with MediaPipe Pose to extract exact spatial vectors of my forearms and overlaying alpha-blended plasma shields. 🔹 The Physics Engine: Implementing vector geometry (dot products) to calculate the exact angle of incidence, allowing autonomous digital projectiles to physically ricochet off my physical arms. 🔹 Extreme Hardware Optimization: Decoupled the neural network inference, the OpenCV rendering loop, and procedural audio synthesis using asynchronous threading to guarantee a flawless 60+ FPS cinematic experience on my RTX 4060 Ti. Generative AI isn't just for chatting; it is rapidly accelerating complex systems engineering. Check out the live inference demo below where I test the EMP trigger! 👇 #ComputerVision #MachineLearning #Python #SoftwareEngineering #OpenCV #GeminiAI #DataScience #Innovation

Computer science should feel accessible, creative, and connected to the world students live in. That’s why I’m excited about imagi and its plug-and-play computer science and AI lessons for grades 3–12. Teachers can introduce Python, AI, and coding fundamentals with confidence, even without prior tech experience. Full lesson plans and materials are provided, AI-assisted debugging helps keep momentum going, and lessons work on Chromebooks, laptops, and tablets. Learn more: https://imagilabs.com/ #education #edtech #STEM #coding #vibecoding #AI #STEMEducation Lovable

Building something I wish existed when I started learning quantum computing. Qiskit Intuition is an open-source interactive lab that teaches quantum computing the way physics should be taught — through visual intuition, not equations first. Drop a gate. Watch the Bloch sphere move. Ask the AI why. It combines a drag-and-drop circuit composer, real-time 3D state visualization, a multi-agent AI tutor (A.C.E.), and a live Qiskit sandbox — all in one place. Still very much a work in progress, but the architecture is taking shape. The latest update adds parameterized gates, an intent-routing AI chat, preset experiments, and a full 6-level curriculum from Python basics to IBM hardware execution. Would love feedback from anyone in quantum education or QC research. 🔗 https://lnkd.in/grQuKrJA #QuantumComputing #Qiskit #OpenSource #MachineLearning #Physics #Quantum

Over the last few weeks, I’ve been working to bridge the gap between human biomechanics and software engineering. Today, I’m incredibly excited to launch my most complex build yet: The Neural-Kinematic-Interface (NKI). 🧠⚡ I built a completely touchless, spatial operating system from scratch using Python, OpenCV, and Deep Learning. Instead of just writing a single script, I engineered a finite-state machine that routes my hand’s skeletal data to power a suite of holographic applications. A few major technical hurdles I had to overcome: 🔹 Spatial 3D Engine: Built a custom projection engine using rotational matrix math to render and manipulate geometry in mid-air (including a procedural DNA double-helix) without relying on heavy game engines. 🔹 The Main Thread Bottleneck: Integrated Asynchronous Threading to run Tesseract OCR and a pyttsx3 Voice AI engine in the background, keeping the real-time computer vision feed at a perfectly smooth frame rate. 🔹 Bullet-Through-Paper Problem: Implemented Continuous Collision Detection (CCD) using vector mathematics to flawlessly detect when my hand intersects with fast-moving digital targets. Exploring how the physics of human motion can directly command digital systems is exactly why I'm passionate about the intersection of AI and software engineering. Check out the demo video below to see the UI in action, and feel free to dive into the full source code and documentation on my GitHub! 🔗 GitHub Repo: https://lnkd.in/gbJDN-xu #Python #ComputerVision #SpatialComputing #DeepLearning #SoftwareEngineering #OpenCV #AI #TechInnovation

Decoding a word is hard work for a six-year-old. So, letting them physically walk through each sound turns an abstract task into something their eyes, ears, and body all learn together. 📚 Code-to-Read from Pitsco pairs systematic, science-of-reading phonics with programmable robots so kids practice sounds with their hands, feet, and brains all firing at once. How it works in the classroom: → Students decode a CVC word, then code a Blue-Bot to walk the sounds across a mat. → Phonics patterns become physical paths kids build, test, and debug together. → Reading, computational thinking, and collaboration grow. Read the full blog: https://lnkd.in/gK_XFJHn #PitscoEducation #EarlyLiteracy #STEMEducation

Every lecture used to begin the same way. “Silence please. Attendance.” And just like that, the first 10 minutes of every class disappeared. What started as a small frustration turned into a question for me and Ansh Raythatha: Why are classrooms in 2026 still dependent on roll calls and spreadsheets? So we built a Contactless Attendance Management System powered by Face Recognition. A faculty member can start a session, upload a group photo or use a live camera feed, and the system automatically marks attendance in real time. At the end of the session, present and absent reports are generated and emailed instantly. What made this project interesting was not just detecting faces — it was making the recognition reliable in real classroom conditions. Behind the scenes: • InsightFace ArcFace generates 512-dimensional facial embeddings for recognition • YOLOv8n-face acts as a secondary detector for crowded group photos • A dual-pass recognition pipeline improves detection coverage in large classrooms • A rolling 5-frame voting mechanism prevents false positives in live mode • Flask powers the backend with role-based Admin and Faculty portals • Dockerized and deployed on Railway with memory optimization under 350MB One of the biggest engineering challenges was accuracy in crowded classrooms. A single frame should not mark someone present just because they briefly appeared in front of the camera. Solving that required threshold tuning, voting logic, embedding validation, and balancing speed with accuracy. The result: • ~95%+ accuracy in live camera mode • ~92% accuracy in 30+ student group photos This project taught me that building AI systems is not only about training models. The real challenge is building systems that people can actually trust and use in real workflows. Live Demo: [https://lnkd.in/gAWDdnNp) GitHub: [https://lnkd.in/gbMjUMEf) Built with Ansh Raythatha. #MachineLearning #ComputerVision #ArtificialIntelligence #Python #Flask #OpenSource #BuildInPublic