Enhanced multi-scale trademark element detection using the improved DETR
Longwen Li, Xiuhui Wang, Wei Qi Yan · 2024
Read Enhanced multi-scale trademark element detection using the improved DETR by Longwen Li, Xiuhui Wang, Wei Qi Yan (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11589202.txt.
Description
Longwen Li, Xiuhui Wang, and Wei Qi Yan teach a detector to pick out the small parts of trademark images in this 2024 paper. Trademark filings keep growing, and so do copycat disputes. Checking images by hand cannot keep pace, and the telling details are often tiny.
Their network builds on a transformer detector and adds two attention tools to the backbone, one that highlights where things are and one that reads the whole scene's context. A feature pyramid then boosts the fine detail that small elements need. On their test set the system reached about ninety-one percent accuracy, ahead of rival detectors. The authors present it as a practical aid for market regulators hunting infringing marks, where missing a small copied element can decide a case.
Catalog record
| CREATOR | Longwen Li, Xiuhui Wang, Wei Qi Yan |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 62 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |