Adventure with Toys Ver.5

“…and they lived happily ever after.”

           The ending I hated the most during my childhood. What happened to Snow White and Prince Charming for three quarters of their life? What happened to Harry Potter after he graduated Hogwarts? Frustrated by the unsatisfactory endings of the movies and novels, I started reconstructing the stories and searching for answers of the numerous ideas popping in my head in an artificial world: a world made of toys.

From the moment I unboxed my first Power Ranger, these little figures have endured a steady presence in my life. My jointed rainbow rangers were not just toys for children; they were a bridge, bringing my imaginations to life. With them on my side, I searched for the answers to epilogue of novels, sudden peculiar ideas, and much more.

The 3-inch soldiers and I became the children of Harry Potter one day, drawing the epilogue of what happened after Voldemort’s death. Another day, we became babysitters of the baby dragons living in the Isle of Berk. After watching Iron man, we became the scientists of stark industry, developing a new suit immune to any physical attacks.

           But soon, I realized that even toys could not satisfy my thirst for imaginations. Even as a king of Middle Earth, I could not find a way to cost-effectively construct the infrastructure for transportation. Even as the master blacksmith of Camelot, I could not logically explain how to repair the Excalibur. And even as Sherlock Homes, I could not persuade the traumatized witness to testify what he saw. I needed something more than the imaginary adventures with my toys: I needed calculations, logic, and reason.

My academic endeavors were the new toys that could finally complete my adventures. From figuring out the Middle Earth construction problem, my interests extended to Union-Find algorithms like Kruskal and Prim, including sorting and grouping edges with the least cost. For more specific analysis, I started pursuing other algorithms; I studied Dijkstra to calculate the actual minimal cost of traveling and researched network flow (a. k. a BP matching) to figure out the maximum number of vehicles on the roads. Through the process of construction, I absorbed the art of mathematics and computational algorithms.

Searching for the appropriate method to repair the Excalibur, I pioneered through deep-learning, understanding how the machines infer unknown shapes and make predictions that human can never make. For a deeper understanding, I started playing with statistics and differential equations, obtaining a solid understanding of feature extraction, classification, and generation. With these new toys, I eventually created my ultimate masterpiece: Fix-GAN, a computed tomography based Generative Adversarial Network model to reconstruct 3D shapes.

           To capture the criminal Moriarty(the villain in Sherlock Homes), I began studying the art of convincing, which led to another major entertainment: debate. Challenging myself in prestigious debate clubs and tournaments, I learned how to structure my speech and deeply analyze my points. For more effective convincing, I took in knowledge on politics and international relations, feeling an extreme thrill when using the comparison of civil society in China and US for a rebuttal in a tournament. When I was confident that I now had the techniques to persuade the witness, I found myself with a powerful toy that I could use in any forms of communication.

           Originating from my childish daydreams, every random adventure with toys offers another motivation, a project idea, or at the very least, a small spark of question that adds a new layer to my academics and how I view the world.

           As my view and perspective enriched, my toys too have stacked layers of memories and knowledge inside them. Holding Power Ranger Red, my oldest and favorite toy, I hear him telling me the stories that once sparked my thoughts. My adventures with toys are the history of my growth.

           And the adventures will never stop, traveling ever after.