PARAMETER LEARNING IN A HIDDEN TRAJECTORY MODEL

Patent №

US 8,010,356

Granted

2011-08-30

Filed 2006

Owner

MICROSOFT CORPORATION

AI components

5

ml · nlp · speech · planning · hardware

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

11356898

Parameters for distributions of a hidden trajectory model including means and variances are estimated using an acoustic likelihood function for observation vectors as an objection function for optimization. The estimation includes only acoustic data and not any intermediate estimate on hidden dynamic variables. Gradient ascent methods can be developed for optimizing the acoustic likelihood function.

AI classification

Speech1.00
Machine learning1.00
AI hardware1.00
Natural language0.97
Planning0.92
Vision0.37
Knowledge representation0.02
Evolutionary computation0.00

Ownership

MICROSOFT CORPORATION

assignment · 173830118

Assignors

DENG, LI, YU, DONG, LI, XIAOLONG, ACERO, ALEJANDRO

On an employer assignment, the assignors are typically the inventors.

© 2026 NYSGPT2525 LLC