Sammenlign metoder
Gennemgå dine valgte metoder side om side; rækker, der afviger, er fremhævet.
| Few-Shot Objektdetektering× | DETR (Detection Transformer)× | |
|---|---|---|
| Fagområde | Dyb læring | Dyb læring |
| Familie | Machine learning | Machine learning |
| Oprindelsesår | 2020 | 2020 |
| Ophavsperson≠ | Xin Wang | Nicolas Carion |
| Type | Neural network architecture | Neural network architecture |
| Oprindelig kilde≠ | Wang, X., Huang, T. E., Darrell, T., Gonzalez, J. E., & Yu, F. (2020). Few-shot object detection with attention-RPN and multi-relation detector. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 9050-9059). link ↗ | Carion, N., Massa, F., Synnaeve, G., Usunier, N., Kirillov, A., & Zagoruyko, S. (2020). End-to-end object detection with transformers. In European Conference on Computer Vision (pp. 213-229). Springer, Cham. DOI ↗ |
| Aliasser | FSOD, Few-shot detection | Detection Transformer, DETR |
| Relaterede≠ | 3 | 4 |
| Resumé≠ | Few-Shot Object Detection (FSOD) is a meta-learning approach that enables detecting novel object classes from only a few annotated examples. Unlike standard object detection requiring hundreds of labeled instances per class, FSOD learns to quickly adapt detection models to new object categories by leveraging knowledge from base categories. | DETR (Detection Transformer) is an end-to-end framework for object detection introduced by Carion et al. in 2020 that reformulates detection as a direct set prediction problem using transformers. Unlike traditional approaches that use hand-crafted post-processing like non-maximum suppression, DETR treats object detection as a sequence-to-sequence problem where the transformer predicts all objects at once. |
| ScholarGateDatasæt ↗ |
|
|