hi everyone !
curently i want to build a voice odering application but i don’t know where to start acually the purpuse of the app is take take user’s speech request then transform it into texte treate it and provide a speech answer in french
So I don’t know what the purpose of a “voice odering app” is but I can tell you would need much more than a speech recognition web app.
For a start, if you want to provide the user with an answer (no matter the language), you need some sort of NLP (Natural Language Processing) tool to understand and classify the user intents and provide the adequate responses.
There many available and your choice is going to vary depending on the amount of skills and time you have and the deployment platform you want to use.
RASA works quite well as a NLP framework for beginners without getting into too much details but it’s also extensively customizable.
We could also mention Dialogflow from Google or the Alexa Skills Kit by Amazon.
The point is you need to choose a platform for your “chatbot” first. STT (and TTS) comes after because integration varies a lot based on the platform you’ve chosen. For example; Dialogflow uses Google’s own models for Speech-To-Text and Text-To-Speech same for Alexa with Amazon’s models. As for RASA, it’s only an NLP framework so you could use something like Coqui-STT and Coqui-TTS for those tasks.
thank you for providing such a clear details
1 Like