QUALCOMM Incorporated (20240331716). LOW-LATENCY NOISE SUPPRESSION simplified abstract
Contents
LOW-LATENCY NOISE SUPPRESSION
Organization Name
Inventor(s)
Jacob Jon Bean of Vista CA (US)
Rogerio Guedes Alves of Macomb Township MI (US)
Vahid Montazeri of Newport Beach CA (US)
Erik Visser of San Diego CA (US)
LOW-LATENCY NOISE SUPPRESSION - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240331716 titled 'LOW-LATENCY NOISE SUPPRESSION
The patent application describes a device with processors that process audio data to suppress noise in audio signals.
- Processors obtain audio data with first and second segments.
- Transform operations are performed on the first segment to generate frequency-domain audio data.
- Machine-learning models use the frequency-domain audio data to generate noise-suppression output.
- Reverse transform operations are performed on the noise-suppression output to generate time-domain filter coefficients.
- Time-domain filtering is done on the second segment using the filter coefficients to produce a noise-suppressed output signal.
Potential Applications: - Audio enhancement in smartphones, tablets, and other electronic devices. - Noise reduction in video conferencing systems. - Improving speech recognition accuracy in voice-controlled devices.
Problems Solved: - Eliminating background noise in audio signals. - Enhancing the quality of audio recordings. - Improving the overall user experience with audio devices.
Benefits: - Clearer audio quality. - Enhanced user satisfaction. - Improved performance of audio processing systems.
Commercial Applications: - Integration into audio editing software for professionals. - Inclusion in noise-canceling headphones for consumers. - Implementation in smart home devices for better voice recognition.
Questions about the technology: 1. How does the device differentiate between noise and desired audio signals for suppression? 2. What are the limitations of the noise-suppression output in terms of audio quality preservation?
Original Abstract Submitted
a device includes one or more processors configured to obtain audio data representing one or more audio signals. the audio data includes a first segment and a second segment subsequent to the first segment. the one or more processors are configured to perform one or more transform operations on the first segment to generate frequency-domain audio data. the one or more processors are configured to provide input data based on the frequency-domain audio data as input to one or more machine-learning models to generate a noise-suppression output. the one or more processors are configured to perform one or more reverse transform operations on the noise-suppression output to generate time-domain filter coefficients. the one or more processors are configured to perform time-domain filtering of the second segment using the time-domain filter coefficients to generate a noise-suppressed output signal.
