High Recognition Accuracy
Combining the respective advantages of the tree model and the neural network model, while ensuring the accuracy of error correction, the recognition accuracy is improved
Based on the tree model and the neural network model, the wrong fragments in the text are identified, the error prompts are given, and the correct suggested text content is given
Combining the respective advantages of the tree model and the neural network model, while ensuring the accuracy of error correction, the recognition accuracy is improved
Not only contains common homonyms, miscellaneous characters and other error correction data, but also accumulates ASR speech recognition easy to recognize errors such as flat tongue, polyphonic characters and other error correction data
Training samples based on error correction data in a variety of scenarios to ensure the stability of the algorithm in complex and changeable scenarios
Integrate multiple resources, provide a knowledge base for semantic search, and provide a comprehensive introduction to search entities
The user will have input errors when searching. The text error correction module can automatically correct the user's search query and provide search results that are more in line with the user's intentions
Embedded text error correction module in the dialogue management system can automatically correct typos in the speech recognition process, effectively improve the accuracy of speech recognition and enhance user experience
