Abstract
One of the challenges of supervised learning training is the need to procure substantial amount of tagged data. A well-known method of solving this problem is to use synthetic data in a copy-paste fashion so that we cut objects and paste them onto relevant backgrounds. Pasting the objects naively results in artifacts that cause models to give poor results on real data. We present a new method for pasting objects on different backgrounds so that the dataset created gives competitive performance on real data by only treating the border of the pasted object using inpainting. We show state-of-the-art results both on instance detection and foreground segmentation.
| Original language | English |
|---|---|
| Title of host publication | Proceedings - 16th International Conference on Signal-Image Technology and Internet-Based Systems, SITIS 2022 |
| Editors | Kokou Yetongnon, Albert Dipanda, Luigi Gallo |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 427-433 |
| Number of pages | 7 |
| ISBN (Electronic) | 9781665464956 |
| DOIs | |
| State | Published - 2022 |
| Event | 16th IEEE International Conference on Signal-Image Technology and Internet-Based Systems, SITIS 2022 - Dijon, France Duration: 19 Oct 2022 → 21 Oct 2022 |
Publication series
| Name | Proceedings - 16th International Conference on Signal-Image Technology and Internet-Based Systems, SITIS 2022 |
|---|
Conference
| Conference | 16th IEEE International Conference on Signal-Image Technology and Internet-Based Systems, SITIS 2022 |
|---|---|
| Country/Territory | France |
| City | Dijon |
| Period | 19/10/22 → 21/10/22 |
Bibliographical note
Publisher Copyright:© 2022 IEEE.
Keywords
- Copy-paste
- Inpainting
- Synthetic data