Qualcomm incorporated (20240104311). HYBRID LANGUAGE TRANSLATION ON MOBILE DEVICES simplified abstract

HYBRID LANGUAGE TRANSLATION ON MOBILE DEVICES

Organization Name

qualcomm incorporated

Inventor(s)

Kee-Hyun Park of San Diego CA (US)

Sungrack Yun of Seongnam (KR)

Kyu Woong Hwang of Daejeon (KR)

HYBRID LANGUAGE TRANSLATION ON MOBILE DEVICES - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240104311 titled 'HYBRID LANGUAGE TRANSLATION ON MOBILE DEVICES

Simplified Explanation

The patent application describes a method for recognizing natural language on a mobile device using a neural network to generate text from audio input and determining the accuracy of the generated text.

• Receiving audio input on a mobile device
• Using a neural network to generate local text from the audio input
• Generating a local confidence value for the accuracy of the local text
• Transmitting data corresponding to the audio input to a remote device
• Receiving remote text and a remote confidence score for accuracy from the remote device
• Outputting the local text if the local confidence value is higher than the remote confidence score, and outputting the remote text if the remote confidence score is higher than the local confidence value

---

1. 1. 1. Potential Applications

- Speech recognition technology - Language translation services - Mobile virtual assistants

1. 1. 1. Problems Solved

- Improving accuracy of speech recognition on mobile devices - Enhancing communication capabilities for users - Streamlining language translation processes

1. 1. 1. Benefits

- Increased efficiency in text generation from audio input - Enhanced user experience with accurate language recognition - Seamless integration of speech-to-text technology on mobile devices

1. 1. 1. Potential Commercial Applications
         1. Optimizing Speech Recognition on Mobile Devices

---

1. 1. 1. Possible Prior Art

One possible prior art for this technology could be the use of neural networks in speech recognition systems on mobile devices. Companies like Google and Apple have been incorporating neural networks in their voice recognition technologies to improve accuracy and performance.

1. 1. 1. Unanswered Questions

1. 1. 1. 1. How does this method handle different accents and dialects in natural language recognition?

The abstract does not specify how the neural network adapts to various accents and dialects in speech recognition.

1. 1. 1. 1. What is the processing time for generating and outputting the text on the mobile device?

The abstract does not mention the speed or efficiency of the method in processing audio input and generating text output.

Original Abstract Submitted

a processor-implemented method for recognizing a natural language on a mobile device includes receiving an audio input. the method further includes using a neural network to generate local text corresponding to the audio input. the method still further includes generating a local confidence value for accuracy of the local text. the method includes transmitting, to a remote device, data corresponding to the audio input. the method further includes receiving remote text corresponding to the data, along with a remote confidence score for accuracy of the remote text. the method still further includes outputting the local text in response to the local confidence value being higher than the remote confidence score, and outputting the remote text in response to the remote confidence score being higher than the local confidence value.