Invention Publication
- Patent Title: Boson Sampler and Neural Network for Data Generation
-
Application No.: US18451255Application Date: 2023-08-17
-
Publication No.: US20240135144A1Publication Date: 2024-04-25
- Inventor: Hugo Wallner , William Clements
- Applicant: ORCA Computing Limited
- Applicant Address: GB London
- Assignee: ORCA Computing Limited
- Current Assignee: ORCA Computing Limited
- Current Assignee Address: GB London
- Priority: GB 15101.3 2022.10.13
- Main IPC: G06N3/045
- IPC: G06N3/045 ; G01N21/45 ; G06N3/0464 ; G06N3/084 ; G06T11/00

Abstract:
Methods are provided for generating a dataset (e.g., an image). According to an example, the method comprises controlling a boson sampler to produce one or more integer sequences, each of the one or more integer sequences representative of a measurement outcome of one or more photodetectors of the boson sampler; determining, from the one or more integer sequences, one or more latent vectors; providing the determined one or more latent vectors to a trained artificial neural network (ANN) configured to convert one or more latent vectors to a generated dataset; and outputting the generated dataset. Methods for training an ANN are also provided. Systems and computer-readable storage media are also described.
Information query