METHOD AND SYSTEM OF SIGNAL DECOMPOSITION USING EXTENDED TIME-FREQUENCY TRANSFORMATIONS

Patent №

US 10,366,705

Granted

2019-07-30

Filed 2017

Owner

Lab

AI components

2

speech · hardware

Assignment

None on record

Dataset

AIPD

2023_r1 edition

Application

15804675

A system and method of decomposing a source signal comprising first and second sound signals, using an extended time-frequency transformation formed by combining a time frequency transformation of a first representation of the source signal with a time-frequency transformation of a second representation of the source signal. The source signal can comprise music, speech, video or other multimedia signals and the decomposition can be controlled by a single knob user interface.

SpeechAI hardwareG10L 21/0272G10L 19/008G10L 21/02G10L 21/0208

AI classification

Speech1.00
AI hardware0.68
Machine learning0.01
Natural language0.00
Evolutionary computation0.00
Planning0.00
Knowledge representation0.00
Vision0.00
© 2026 NYSGPT2525 LLC