Generating Virtual Reality Content Based on Corrections to Stitching Errors

Patent №

US 11,348,202

Granted

2022-05-31

Filed 2020

Owner

JAUNT INC.

+1 more

Lab

AI components

2

vision · evo

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

17084408

Illustrative methods and systems for generating virtual reality content based on corrections to stitching errors are described. An illustrative computer-implemented method includes receiving raw virtual reality video data representing images recorded by camera modules of a camera array, using a stitching algorithm to stitch the images together to generate an initial virtual reality render, determining that the initial virtual reality render has a stitching error, determining a correction for the stitching error, and modifying the stitching algorithm based on the correction for the stitching error. The modified stitching algorithm may be used to generate a corrected virtual reality render that corrects the stitching error.

VisionEvolutionary computationG06T 3/4038G06T 1/20H04N 13/111H04N 13/246H04N 23/45H04N 23/661H04N 23/698H04N 25/41+3 more

AI classification

Vision1.00
Evolutionary computation0.87
Knowledge representation0.26
Planning0.02
Natural language0.01
Machine learning0.01
AI hardware0.01
Speech0.00

Ownership

JAUNT INC.

assignment · 542160866

VERIZON PATENT AND LICENSING INC.

assignment · 542160877

Assignors

VAN HOFF, ARTHUR

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

© 2026 NYSGPT2525 LLC