Koushik Dey Sarkar

What if a bird could learn to fly through pipes — on its own? No hardcoded logic, no cheat sheets, just raw trial and error. That’s what I set out to build: a self-learning version of Flappy Bird that runs entirely in your browser, trains in real-time, and visually evolves right before your eyes. Welcome to my first major reinforcement learning project — Flappy Bird AI.

“If I’m an advocate for anything, it’s to move. As far as you can, as much as you can. Across the ocean, or simply across the river.” — Anthony Bourdain The world is full of icons — some gilded in gold, others carved in stone — but perhaps none more ubiquitous, more quietly democratic, than the humble white plastic chair. You’ve seen it. We all have. At a street café in Paris. A family gathering in Manila. A roadside tea stall in Delhi. A beachside barbecue in Rio. A market in Marrakech.

Koushik Sarkar · Follow 4 min read · -- Unlocking Computer Vision: CNNs Explained Simply CNN Diagram Introduction: Seeing Like a Computer Ever wonder how computers “see” and recognize images? Imagine recognizing a friend. Your eyes detect shapes, your brain combines them into a face, and you say, “That’s Ryan!” This is similar to how a Convolutional Neural Network (CNN) works, but in a digital way. CNNs are the powerhouse behind image recognition, object detection, and more.