Tencent Cloud Launches ASR Speech Recognition System with Large Model Deployment
-
Tencent Cloud ASR (Automatic Speech Recognition) is a speech recognition system launched by Tencent Cloud. After the latest upgrade, Tencent Cloud ASR can better handle dialects and noise, improving recognition accuracy and comprehension capabilities.
In terms of dialects, Tencent Cloud ASR can recognize 23 local dialects and handle mixed dialect scenarios. Through its self-developed multilingual hybrid recognition engine, Tencent Cloud ASR can quickly identify and seamlessly switch between different dialects, all with a single engine. Using distillation algorithms and semi-supervised training, Tencent Cloud ASR efficiently learns from both labeled and unlabeled data, even processing minor dialects of ethnic minorities.
In terms of semantic recognition, Tencent Cloud ASR utilizes large language models (LLM) to extract intent through contextual semantics and big data, thereby improving recognition accuracy. For words with similar pronunciations, Tencent Cloud ASR can still select the correct one. Additionally, AI can more accurately recognize special scenarios or professional terminology.
Regarding noise filtering, Tencent Cloud ASR has developed anti-interference capabilities through self-learning and optimization algorithms. It can automatically filter out noise to enhance recognition precision.
Tencent Cloud ASR has been widely applied in both Tencent's internal products and various external industries, covering multiple scenarios such as call quality inspection, real-time meeting transcription, and voice input methods. The product handles billions of daily API calls and serves thousands of enterprise clients, both internal and external.
Experience URL: https://cloud.tencent.com/product/asr