VoiceTalk: A No-Code Approach for Creating Voice-Controlled Smart Home Applications
Researchers from the College of Artificial Intelligence have developed VoiceTalk, a no-code approach to creating voice-controlled smart home applications without requiring programming expertise. This innovative platform utilizes IoTtalk, an IoT application development platform for managing a diverse range of IoT devices. The research was funded by the National Science And Technology Council, Ncku Miin Wu School of Computing, Quanta And Nycu/ncku Ai Joint Research Center, China Medical University Hospital, and Research Center For Information Technology Innovation, Academia Sinica.
Key Takeaways:
- VoiceTalk is a no-code approach that simplifies the development of Google Home-like applications, integrating IoTtalk with Google Home.
- The platform leverages its understanding of smart appliances in the room/house to generate specific prompts.
- VoiceTalk significantly reduces voice-to-text translation errors, from 5.13% to 0.54% for the Web Speech API and from 2.25% to zero for Whisper Medium.
- The research compared the translation accuracy of 7 Automatic Speech Recognition (ASR) systems and made two contributions: simplifying the development of Google Home-like applications and reducing voice-to-text translation errors.
- The Device LLM Agent of VoiceTalk can be easily extended to integrate IoTtalk with other voice platforms, such as AWS Alexa.
- The research concluded that for small-sized open-source LLMs such as Llama 3.2 3B, the errors are reduced to 0.72% for the Web Speech API and to zero for Whisper Medium.
Statistics:
- 5.13% to 0.54% reduction in voice-to-text translation errors for the Web Speech API
- 2.25% to zero reduction in voice-to-text translation errors for Whisper Medium
- 0.72% to zero reduction in voice-to-text translation errors for small-sized open-source LLMs such as Llama 3.2 3B
- 7 Automatic Speech Recognition (ASR) systems compared in the research
- 2 contributions made by the research: simplifying the development of Google Home-like applications and reducing voice-to-text translation errors
Sources:
- VoiceTalk: A No-Code Approach for Creating Voice-Controlled Smart Home Applications. IEEE Open Journal of the Computer Society, 2025, 6():874-883
- doi-org.sdpl.idm.oclc.org/10.1109/OJCS.2025.3576725
- NewsRx. Research Results from College of Artificial Intelligence Update Knowledge of Computer Science (VoiceTalk: A No-Code Approach for Creating Voice-Controlled Smart Home Applications). Computer Weekly News. July 16, 2025; p 548