Omnivision Technologies, Inc.

Zoom Communications, Inc.

Cheng-Lun Hu

Dong Shi

Chung-An Wang

Lu Chang

Zhaofeng Jia

Qiyong Liu

Qi Yang

Jiachuan Deng

Online audio and video conference applications can utilize a noise removal module to eliminate unwanted audio from a participant's speech. A noise removal module can rely on differentiating between human speech versus other audio to filter out noise. However, in some conference environments, participant and non-participant human speech can be present. Artificial intelligence models can be trained to detect both noise and non-participant audio, based on a variety of factors. The models can label captured audio and various noise removal modules can filter noise based on the output of the models.

Dynamic noise and speech removal

A surveillance system includes an infrared sensor system coupled to output an infrared signal in response to receiving infrared light, and an audio recording system coupled to output an audio signal in response to recording sound. An image sensor is system coupled to output an image signal in response to receiving image light. A controller is coupled to the infrared sensor system, the audio recording system, and the image sensor system. The controller includes logic that when executed by the controller causes the surveillance system to perform operations including receiving the infrared signal from the infrared sensor system, activating the audio recording system to record the sound, and activating the image sensor system to output the image signal.

Reducing false alarms in surveillance systems

An electronic device to equalize sound includes a microphone coupled to receive the sound and output sound data. The sound has amplitude that changes with respect to time, and the sound includes one or more frequencies. The sound data indexes the amplitude with respect to the time. The electronic device also includes a controller coupled to microphone, and the controller includes logic that when executed by the controller causes the electronic device to perform operations. The operations may include: receiving the sound data from the microphone with the controller; adjusting the amplitude of the sound included in the sound data across the one or more frequencies using a filter disposed in the logic; and outputting filtered sound data.

Low complexity loudness equalization

A speaker system featuring automatic volume adjustment to avoid excess distortion includes a first converter for converting an input signal to an input frequency domain signal; and a second converter for converting a microphone signal to an environmental frequency domain signal. A distortion analyzer receives output of the first and second converter and determines at least one distortion level and provides at least one gain adjustment signal used to control a corrector to adjust gain of a speaker driver. The system performs a method of reducing distortion by automatically adjusting volume in each of a plurality of frequency bands including converting an input signal to a frequency domain input signal; converting a microphone signal to an environmental frequency domain signal; determining a distortion level; generating a gain adjustment signal; and adjusting gain of a signal provided to a speaker driver.

Growing community of inventors

Singapore, Singapore

Cheng-Lun Hu