Speech Enhancement using Deep Learning Techniques

In current era of digital technology, speech enhancement has become one of the important challenges in day-to-day life for smoother conversation and understanding. Most of the recorded or live speech audio signals contain some form of noise such as street noise, barking sounds of dogs, conversation between multiple people, construction noise, car honking etc. The key feature of speech enhancement system is to execute in real time on a recorded waveform with minimal possible lag on any localized hardware with minimum computational resource.