Method and system for generating adversarial attacks on video recognition models
US11776241B1 · kind B1 · utility
Assignee
Inventors
Key dates
| Filing date | Sep 30, 2022 |
| Grant date | Oct 3, 2023 |
| Priority date | — |
| Expiry date | Sep 30, 2042 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06V20/40
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
A method for generating adversarial attacks on a video recognition model is disclosed, including (a) generating the content of bullet-screen comments (BSCs) by an image captioning model for a clean video sample; (b) generating an adversarial video sample by inserting the BSCs into the clean video sample; (c) using the adversarial video sample to attack the video recognition model; (d) receiving rewards from the environment; (e) optimizing the position and transparency of BSCs by an reinforced learning (RL) agent according to the received rewards; (f) updating the adversarial video sample and using it to attack the video recognition model; and iteratively repeating steps of (d)-(f), until a predefined condition is matched.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.