Tiến Dũng Email & Phone Number

For me, the first job is to learn, not just to "work." I don't care much about salary, as long as it's enough to keep me alive. I want a job challenging enough so that I can work and learn optimally. My curiosity is boundless, and I'm always eager to learn.If you want an employee who is highly skilled from the start, don't hire me. The last thing I want is to hate my job. However, if you're looking for someone who is conscientious and always trying to improve, that's me!Most of my experience in Machine Learning, Game Development, and Virtual Reality application Development comes from school projects. However, due to the broad interests of my trajectory (Master SIIA, UBO), most of what taught in school is at a surface level. Therefore, much of my knowledge in these three domains is self-taught. Since I transitioned mid-way from biotechnology to computer science, I can't say I have a solid base in the latter. That's why I'm constantly learning to build a solid foundation.I am a firm believer in the First Principles thinking, which is the practice of decomposing things down to the fundamental axioms, before reasoning up from only relevant ones. I truly believe that with a deep understanding of what I'm doing, I can solve any problem.

• 

  Deep Learning Researcher

  Centre Européen De Réalité Virtuelle

  Feb 2024 - Present - Brest, Brittany, France

  Neural Architecture Search for Efficient Neural Networks:My internship focused on addressing the increasing need for efficient neural network models, particularly in terms of energy consumption and computational efficiency. Neural Architecture Search (NAS) is a sub-field of AutoML that includes methods to automatically search for neural architectures meeting specific requirements, typically identifying high-performance networks for a specific task. This internship aimed to develop a new NAS method that finds neural network models with a good performance/efficiency tradeoff.Key achievements:- Successfully proposed a new NAS method that rapidly finds neural network architectures with both high performance and improved efficiency, without needing to train candidate architectures.- Inspired by the brain criticality theory, developed a new metric to assess a model's efficiency.