METHOD AND DEVICE FOR RECONSTRUCTING ACOUSTIC DATA AND ANIMATION DATA IN SYNCHRONIZATION

Patent №

US 6,735,738

Granted

2004-05-11

Filed 1999

Owner

FUJITSU LIMITED

Lab

AI components

1

speech

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

09421226

A pair of acoustic data and animation data that are in synchronization are compressed or expanded, and reconstructed in synchronization in the following manner. The acoustic data and the animation data are divided into a plurality of segments. A parameter time series that determines the degree of compression or expansion of the acoustic data and the animation data is determined for each divided segment. The divided acoustic data is further divided at equal intervals into input frames. An output frame length is determined based on the determined parameter time series. A correspondence table for synchronous points between the acoustic data before compression or expansion and the acoustic data after compression or expansion with respect to the acoustic data to be compressed or expanded is output based on the determined parameter time series. The acoustic data is compressed or expanded based on the correspondence table. The animation data is compressed or expanded based on the correspondence table.

SpeechG11B 20/00007H04N 9/8063H04N 5/781H04N 5/85H04N 5/907H04N 9/8042

AI classification

Speech0.90
Vision0.44
AI hardware0.07
Natural language0.02
Knowledge representation0.00
Machine learning0.00
Planning0.00
Evolutionary computation0.00

Ownership

FUJITSU LIMITED

assignment · 103390492

Assignors

KOJIMA, HIDEKI

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

© 2026 NYSGPT2525 LLC