I recently attempted to install OpenVoiceOS/MyCroft on my OrangePi PC and found the process to be quite daunting. At the end of the day I didn't manage to install, no meaningful error in the log. The installation is riddled with numerous dependencies and compiling everything on an armv7 single board computer is time-intensive, taking hours.
My goal is straightforward: to create a basic voice assistant capable of providing weather updates, financial data, and utilizing a Large Language Model (LLM) for answering various questions. I've already constructed a prototype that leverages Microsoft Speech Services, which you can check out here: https://github.com/dzianisv/AssistantPlato.
Unfortunately, Microsoft's Speech Services don't offer support for armv7, so I'm considering a transition to OpenAI's Whisper. However, a significant hurdle remains: I need to develop a reliable voice activation system. I'm leaning towards using PocketSphinx for this purpose, but I'm open to suggestions.
Has anyone tackled similar challenges or have experience with voice activation on limited-resource devices? Any insights or advice would be greatly appreciated.
I recently attempted to install OpenVoiceOS/MyCroft on my OrangePi PC and found the process to be quite daunting. At the end of the day I didn't manage to install, no meaningful error in the log. The installation is riddled with numerous dependencies and compiling everything on an armv7 single board computer is time-intensive, taking hours.
My goal is straightforward: to create a basic voice assistant capable of providing weather updates, financial data, and utilizing a Large Language Model (LLM) for answering various questions. I've already constructed a prototype that leverages Microsoft Speech Services, which you can check out here: https://github.com/dzianisv/AssistantPlato. Unfortunately, Microsoft's Speech Services don't offer support for armv7, so I'm considering a transition to OpenAI's Whisper. However, a significant hurdle remains: I need to develop a reliable voice activation system. I'm leaning towards using PocketSphinx for this purpose, but I'm open to suggestions.
Has anyone tackled similar challenges or have experience with voice activation on limited-resource devices? Any insights or advice would be greatly appreciated.