Lip-to-Speech
I spent time researching lip-to-speech generation—from data processing and model training to VSR, ASR guidance, diffusion models, and extensive experimental validation. Many ideas were tested; quite a few were discarded.
于喧嚣之外,写自己的答案
Xi'an Polytechnic University · B.Eng. student in Cyberspace Security
Exploring the space between security, AI, and systems. This is where I keep things I have built, lessons I have learned, and questions I have not fully answered yet.
I am 0xWay1, an undergraduate in Cyberspace Security at Xi'an Polytechnic University, currently focused on cybersecurity, artificial intelligence, and AI infrastructure. I enjoy working with code and servers, reading papers, and running experiments—slowly making sense of the questions that genuinely interest me. More often than reaching an answer quickly, I enjoy thinking things through until a problem begins to reveal its original shape.
Over time, I have become less eager to draw conclusions. Some questions only become clear when you move closer; some answers need time before their outlines appear. Directions change and ideas evolve, but anything given real time and thought leaves a trace somewhere.
That is why this site exists. It records projects I have built, things I have read, mistakes I have made, and thoughts I have not yet fully resolved. Looking back one day, I may find my earlier self a little inexperienced—and that is fine. Being able to see how an idea changed, little by little, is itself the meaning of keeping a record.
“I'm going to live every minute of it.”
— Soul (2020)
I cannot quite say where the future will lead.
But at least, I hope no part of the journey is hurried past unnoticed. ( ̄▽ ̄)
Latest posts
I spent time researching lip-to-speech generation—from data processing and model training to VSR, ASR guidance, diffusion models, and extensive experimental validation. Many ideas were tested; quite a few were discarded.
I am exploring how AI agents can work with cybersecurity, with a current focus on vulnerability intelligence, security analysis, automation, and the permission and safety boundaries of agents in real security environments.
I work mainly with Python and C/C++, with an emphasis on deep learning, cybersecurity, and systems. These are the technologies I use regularly or am currently learning.