If you look for speech recognition examples for microcontrollers, you almost always end up with wake word detection or keyword spotting: the micro_speech example of TensorFlow Lite Micro, which recognizes “yes” and “no”; Espressif’s ESP-SR with WakeNet for the wake word and MultiNet for a list of commands; keyword spotting Read more
Classic embedded text-to-speech (formant synthesizers like SAM or eSpeak) is tiny and fast, but it sounds like a robot from the 80s. Neural TTS sounds natural, but it is usually something you run on a PC or in the cloud. I wanted to know: what is the cheapest hardware that Read more
To produce the most natural-sounding speech, modern Text-To-Speach (TTS) systems rely on neural networks. TinyTTS is a proof of concept: It asks whether a VITS-style neural TTS model can run on a microcontroller at all. It is a header-only C++ port of tiny-tts. The model has about 1.6M parameters. It Read more
There are quite a few popular TTS libraries for Arduino but most of them suffer from the same problem: The TTS function is not properly separated from the output function. Architechture In a properly architected solution we would have a A TTS function which produces platfrom independent PCM data An Read more
MBROLA is a speech synthesizer based on the concatenation of diphones. It takes a list of phonemes as input, together with prosodic information (duration of phonemes and a piecewise linear description of pitch), and produces speech samples on 16 bits (linear), at the sampling frequency of the diphone database. Today Read more
It was quite a challenge to convert eSpeak NG to an Arduino Library. In the meantime I managed to publish a first round of error corrections. The eSpeak NG is a compact open source software text-to-speech synthesizer for Linux, Windows, Android and other operating systems. It supports more than 100 Read more
It was quite a challenge to convert eSpeak NG to an Arduino Library. The eSpeak NG is a compact open source software text-to-speech synthesizer for Linux, Windows, Android and other operating systems. It supports more than 100 languages and accents. It is based on the eSpeak engine created by Jonathan Read more
I converted eSpeak NG to an Arduino Library. The eSpeak NG is a compact open source software text-to-speech synthesizer for Linux, Windows, Android and other operating systems. It supports more than 100 languages and accents. It is based on the eSpeak engine created by Jonathan Duddington. eSpeak NG uses a Read more
In my introductory blog, I have described that it would be cool to remove the silence from the recorded samples. If we look at one of the recordings we can see that they start and end with some silence: I have created a SilenceRemovalConverter which does exactly this. The first Read more
I the last Blog I introduced you to my new Simple TTS Arduino Library which is available on Github. I used the Text-to-Speech (TTS) functionality of Google Translate to generate the mp3 files, which was quite tedious to do manually. Then I executed xxd to generate a .h file from Read more