20240038254. SIGNAL PROCESSING DEVICE, SIGNAL PROCESSING METHOD, SIGNAL PROCESSING PROGRAM, LEARNING DEVICE, LEARNING METHOD, AND LEARNING PROGRAM simplified abstract (Nippon Telegraph and Telephone Corporation)

SIGNAL PROCESSING DEVICE, SIGNAL PROCESSING METHOD, SIGNAL PROCESSING PROGRAM, LEARNING DEVICE, LEARNING METHOD, AND LEARNING PROGRAM

Organization Name

Nippon Telegraph and Telephone Corporation

Inventor(s)

Tsubasa Ochiai of Musashino-shi, Tokyo (JP)

Marc Delcroix of Musashino-shi, Tokyo (JP)

Yuma Koizumi of Musashino-shi, Tokyo (JP)

Hiroaki Ito of Musashino-shi, Tokyo (JP)

Keisuke Kinoshita of Musashino-shi, Tokyo (JP)

Shoko Araki of Musashino-shi, Tokyo (JP)

SIGNAL PROCESSING DEVICE, SIGNAL PROCESSING METHOD, SIGNAL PROCESSING PROGRAM, LEARNING DEVICE, LEARNING METHOD, AND LEARNING PROGRAM - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240038254 titled 'SIGNAL PROCESSING DEVICE, SIGNAL PROCESSING METHOD, SIGNAL PROCESSING PROGRAM, LEARNING DEVICE, LEARNING METHOD, AND LEARNING PROGRAM

Simplified Explanation

The abstract describes a signal processing device that uses neural network technology to extract a specific audio class from a mixture audio signal. The device receives extraction target information indicating the desired audio class and uses a feature value of the mixture audio signal to extract the audio signal of the indicated class.

• The device is designed to process audio signals and extract specific classes from a mixture signal.
• It uses a neural network to perform the extraction based on the extraction target information and feature value of the mixture signal.
• The extraction target information indicates which audio class is to be extracted.
• The device outputs the result of the extraction, providing the audio signal of the indicated class from the mixture signal.

Potential Applications:

• Audio classification and separation: This technology can be used in various applications where it is necessary to extract specific audio classes from a mixture signal, such as in speech recognition systems or audio editing software.
• Noise cancellation: By extracting specific audio classes, such as background noise, this technology can be used to enhance the quality of audio recordings or improve speech intelligibility in noisy environments.

Problems Solved:

• Efficient audio extraction: The device solves the problem of extracting specific audio classes from a mixture signal, which can be challenging due to overlapping frequencies and complex audio environments.
• Automation: The use of neural networks allows for automated extraction based on the extraction target information, reducing the need for manual intervention.

Benefits:

• Improved audio processing: By extracting specific audio classes, the device can enhance the quality and intelligibility of audio signals.
• Versatility: The device can be applied to various audio processing tasks, providing flexibility in different applications.
• Efficiency: The use of neural networks enables efficient and automated extraction, saving time and effort compared to manual methods.

Original Abstract Submitted

a signal processing device includes processing circuitry configured to receive an input of extraction target information indicating which audio class of an audio signal is to be extracted from a mixture audio signal constituted by a mixture of audio signals of a plurality of audio classes, and output a result of extracting the audio signal of the audio class indicated by the extraction target information from the mixture audio signal, with a neural network by using a feature value of the mixture audio signal and the extraction target information.