Patent · US Active

Multi-scale transformer for image analysis

US12217382B2 · kind B2 · utility

0Cited by
0References
22Claims
0Family size

Assignee

Inventors

Key dates

Filing dateDec 4, 2023
Grant dateFeb 4, 2025
Priority date
Expiry dateDec 4, 2043

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06T2207/30168
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

The technology employs a patch-based multi-scale Transformer (300) that is usable with various imaging applications. This avoids constraints on image fixed input size and predicts the quality effectively on a native resolution image. A native resolution image (304) is transformed into a multi-scale representation (302), enabling the Transformer's self-attention mechanism to capture information on both fine-grained detailed patches and coarse-grained global patches. Spatial embedding (316) is employed to map patch positions to a fixed grid, in which patch locations at each scale are hashed to the same grid. A separate scale embedding (318) is employed to distinguish patches coming from different scales in the multiscale representation. Self-attention (508) is performed to create a final image representation. In some instances, prior to performing self-attention, the system may prepend a learnable classification token (322) to the set of input tokens.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.