Intro

Introduction of this Blog MLSP.ai.

Machine Learning (ML) is not something new, it has been around almost since computer were invented. As computing advances and new architectures emerge, ML evolves too. In the recent years, Deep Learning, a subset of ML, has shown so much promise that it had become key driver of computing roadmap.

Starting as a DSP engineer, the journey from coding FFT in assembly to applying multi-layer LSTM doing noise reduction has been adventureful, requiring many neural reorganization in my brain. Just this subject of how ML differs to conventional approach includes a wealth of topics, which I feel would be wasteful if they only live inside Jira tickets.

This Blog will have an emphasis on the Signal Processing application of ML, hence the name MLSP. Worth highlighting, ML deployment, especially on small footprint embedded systems, will be covered in extend as well. It’s an area I found much less mentioned than model & algorithm development.

“Having it running in your hand is much more fun, isn’t it 😉”