On Thu, Apr 22, 2021, 9:46 AM stefan.reich.maker.of.eye via AGI < agi@agi.topicbox.com> wrote:
> I always thought that an FFT is the most likely first step in speech > recognition, seeing as you can almost recognize it visually from a 2D > frequency/time plot as a human. > Sort of. A FFT requires dividing up the audio samples into blocks. The cochlea instead applies bandpass filters to produce an array of continuous signals. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Td08f1cb9cdd5e5d9-M3c6e1923b8173c0c099239b2 Delivery options: https://agi.topicbox.com/groups/agi/subscription
