We make Typecast AI, AI powered content creation service powered by emotional voice AI and avatar. Visit http://typecast.ai for more details

I was responsible for leading machine learning algorithm for always-on microphone. I developed and commercialized voice activation, sound event detection, audio scene classification, contents recognition/matching, etc with team members.I hold more than 40 US patents including granted and pending.The list of my papers and patents are found at Google scholarhttps://scholar.google.co.kr/citations?user=67KT8ZAAAAAJ&hl=en

I worked on speech and audio signal processing algorithm development based on statistical data-driven methods such as blind source separation, independent component analysis, non-negative matrix factorization, and other machine learning algorithms. Accomplishments: 1. Developed dual mic noise reduction algorithm, applied to commercial products. 2. Developed dynamic range control, automatic gain control, and acoustic echo canceler for stable voice call. 3. Developed loudness compensation algorithm for Rx voice signal under noisy environment based on human auditory model. 4. Filed several patents and wrote academic papers on audio signal processing.

I worked as a research staff in the area of machine learning and signal processing at Institute for Neural Computation.Accomplishments: 1. I developed a new algorithm on blind source separation and sound localization. 2. I wrote several academic papers on machine learning and graphical model approach to audio signal processing. 3. I filed patents and transfer technology and software to UCSD intellectual property department.