Invention Grant
- Patent Title: High-fidelity generative image compression
-
Application No.: US17107684Application Date: 2020-11-30
-
Publication No.: US11750848B2Publication Date: 2023-09-05
- Inventor: George Dan Toderici , Fabian Julius Mentzer , Eirikur Thor Agustsson , Michael Tobias Tschannen
- Applicant: Google LLC
- Applicant Address: US CA Mountain View
- Assignee: Google LLC
- Current Assignee: Google LLC
- Current Assignee Address: US CA Mountain View
- Agency: Fish & Richardson P.C.
- Main IPC: G06V10/00
- IPC: G06V10/00 ; H04N19/91 ; H04N19/124 ; G06N3/088 ; H04N19/154 ; G06N3/045

Abstract:
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an encoder neural network configured to receive a data item and to process the data item to output a compressed representation of the data item. In one aspect, a method includes, for each training data item: processing the data item using the encoder neural network to generate a latent representation of the training data item; processing the latent representation using a hyper-encoder neural network to determine a conditional entropy model; generating a compressed representation of the training data item; processing the compressed representation using a decoder neural network to generate a reconstruction of the training data item; processing the reconstruction of the training data item using a discriminator neural network to generate a discriminator network output; evaluating a first loss function; and determining an update to the current values of the encoder network parameters.
Public/Granted literature
- US20220174328A1 High-Fidelity Generative Image Compression Public/Granted day:2022-06-02
Information query