The WeChat team has issued a statement concerning the recently introduced 'WeChat Dialect Initiative' mini-program. Designed to refine dialect speech recognition capabilities, this mini-program encourages users to contribute dialect recordings, a concept sparked by user input highlighting the inadequate support for various dialects. In the nascent stages of the initiative, certain imperfections emerged in the dialect sentences produced by the system. Subsequently, the technical team undertook a thorough review and cleanup of the language corpus, instituting a rigorous review protocol to ensure quality.

Regarding privacy safeguards, data acquisition strictly adheres to the tenets of legality, legitimacy, and necessity, and is exclusively employed for pivotal enhancements such as model training. This data is neither disclosed to third parties nor utilized for alternative commercial ventures, with desensitization techniques and other security protocols firmly in place.

This update also places a premium on elevating the user experience, with aspirations to augment the language corpus, encompass a broader spectrum of dialects, and elevate recognition precision moving forward. The system will propose dialects tailored to the user's locale and incorporate features like voice playback and self-deletion options. Presently, the mini-program has initiated recording avenues for an array of dialects, including Cantonese and Wu. WeChat advises users to ensure a stable internet connection and record in tranquil settings. A dedicated channel has been established to promptly address technical queries raised by users, with responses promised within 48 hours.