公开(公告)号：US11527019B2

公开(公告)日：2022-12-13

申请号：US16875884

申请日：2020-05-15

Applicant: Amazon Technologies, Inc.

Inventor： Luitpold Staudigl ,  Thomas Sydney Austin Wallis ,  Mike Mueller ,  Muhammad Bilal Javed ,  Pablo Barbachano

IPC: G06T9/00 ,  G06T3/40 ,  G06T7/00 ,  G06N3/08 ,  G06N20/20 ,  G06K9/62

Abstract: One or more multi-stage optimization iterations are performed with respect to a compression algorithm. A given iteration comprises a first stage in which hyper-parameters of a perceptual quality algorithm are tuned independently of the compression algorithm. A second stage of the iteration comprises tuning hyper-parameters of the compression algorithm using a set of perceptual quality scores generated by the tuned perceptual quality algorithm. The final stage of the iteration comprises performing a compression quality evaluation test on the tuned compression algorithm.

公开(公告)号：US20210358178A1

公开(公告)日：2021-11-18

申请号：US16875884

申请日：2020-05-15

Applicant: Amazon Technologies, Inc.

Inventor： Luitpold Staudigl ,  Thomas Sydney Austin Wallis ,  Mike Mueller ,  Muhammad Bilal Javed ,  Pablo Barbachano

IPC: G06T9/00 ,  G06T3/40 ,  G06T7/00 ,  G06N3/08 ,  G06N20/20 ,  G06K9/62

Abstract: One or more multi-stage optimization iterations are performed with respect to a compression algorithm. A given iteration comprises a first stage in which hyper-parameters of a perceptual quality algorithm are tuned independently of the compression algorithm. A second stage of the iteration comprises tuning hyper-parameters of the compression algorithm using a set of perceptual quality scores generated by the tuned perceptual quality algorithm. The final stage of the iteration comprises performing a compression quality evaluation test on the tuned compression algorithm.

公开(公告)号：US12002261B2

公开(公告)日：2024-06-04

申请号：US18064192

申请日：2022-12-09

Applicant: Amazon Technologies, Inc.

Inventor： Luitpold Staudigl ,  Thomas Sydney Austin Wallis ,  Mike Mueller ,  Muhammad Bilal Javed ,  Pablo Barbachano

IPC: G06V10/82 ,  G06F18/214 ,  G06F18/25 ,  G06N3/082 ,  G06N3/086 ,  G06N20/20 ,  G06T3/4046 ,  G06T7/00 ,  G06T9/00 ,  G06V10/774 ,  G06V10/778 ,  G06V10/80

CPC classification number: G06V10/82 ,  G06F18/214 ,  G06F18/251 ,  G06N3/082 ,  G06N3/086 ,  G06N20/20 ,  G06T3/4046 ,  G06T7/0002 ,  G06T9/002 ,  G06V10/774 ,  G06V10/7784 ,  G06V10/7788 ,  G06V10/803 ,  G06T2207/30168

Abstract: One or more multi-stage optimization iterations are performed with respect to a compression algorithm. A given iteration comprises a first stage in which hyper-parameters of a perceptual quality algorithm are tuned independently of the compression algorithm. A second stage of the iteration comprises tuning hyper-parameters of the compression algorithm using a set of perceptual quality scores generated by the tuned perceptual quality algorithm. The final stage of the iteration comprises performing a compression quality evaluation test on the tuned compression algorithm.

公开(公告)号：US11544562B2

公开(公告)日：2023-01-03

申请号：US16875887

申请日：2020-05-15

Applicant: Amazon Technologies, Inc.

Inventor： Thomas Sydney Austin Wallis ,  Luitpold Staudigl ,  Muhammad Bilal Javed ,  Pablo Barbachano ,  Mike Mueller

IPC: G06N3/08 ,  G06N3/04 ,  G06F30/27

Abstract: Respective labels indicative of compression-related quality degradation for a set of media object tuples which meet a divergence criterion are obtained; each tuple comprises a reference media object and a pair of corresponding compressed media object versions. Pairs of training records for a machine learning model are generated using the labeled media object tuples and multiple perceptual quality algorithms, with each training record comprising respective perceived quality degradation scores generated by each of the multiple algorithms for a given compressed media object of a tuple. A machine learning model is trained, using the record pairs, to predict quality degradation scores for compressed media objects.

公开(公告)号：US20210357745A1

公开(公告)日：2021-11-18

申请号：US16875887

申请日：2020-05-15

Applicant: Amazon Technologies, Inc.

Inventor： Thomas Sydney Austin Wallis ,  Luitpold Staudigl ,  Muhammad Bilal Javed ,  Pablo Barbachano ,  Mike Mueller

IPC: G06N3/08 ,  G06N3/04

Abstract: Respective labels indicative of compression-related quality degradation for a set of media object tuples which meet a divergence criterion are obtained; each tuple comprises a reference media object and a pair of corresponding compressed media object versions. Pairs of training records for a machine learning model are generated using the labeled media object tuples and multiple perceptual quality algorithms, with each training record comprising respective perceived quality degradation scores generated by each of the multiple algorithms for a given compressed media object of a tuple. A machine learning model is trained, using the record pairs, to predict quality degradation scores for compressed media objects.