voice-driven-npc

To add voice interaction in UrhoX Lua, this framework integrates HTTP endpoints for ASR and TTS services with local intent parsing to extract user commands and generate text or speech NPC responses. It handles both voice input recognition and speech output.

Yes, you can use local rules for intent parsing without an LLM. The framework supports extracting user intent and entities locally, while also offering the option to integrate an LLM for more complex voice command parsing if needed.

Yes, network access and HTTP client capabilities are required to run voice-driven NPCs. The framework uses HTTP endpoints to communicate with external ASR and TTS services for processing voice input and generating speech output.

The best way to initialize a voice-activated NPC is using the skill to set up a character, such as a 'Merchant', with specific trigger keywords like 'buy', 'sell', 'greet', and 'ask_info' to route voice commands.

Yes, voice-driven NPC interaction supports both text and speech responses. After processing voice input through ASR and intent parsing, the framework can generate outputs using TTS services or display them as text.

You should not use HTTP endpoints for ASR and TTS if your UrhoX Lua game environment lacks network access or HTTP client capabilities, as this framework explicitly requires external service connectivity to function.