Comparer des méthodes

	Apprentissage par transfert avec segmentation d'instances	Apprentissage par transfert avec détection d'objets
Domaine	Apprentissage profond	Apprentissage profond
Famille	Machine learning	Machine learning
Année d'origine	2017 (Mask R-CNN); transfer learning paradigm: 2010	2010–2014
Auteur d'origine	He, K. et al. (Mask R-CNN); transfer learning framework: Pan & Yang	Girshick, R. et al. (R-CNN line); Pan & Yang (transfer learning framework)
Type	Transfer learning applied to instance segmentation	Transfer learning / fine-tuning
Source fondatrice	He, K., Gkioxari, G., Dollar, P., & Girshick, R. (2017). Mask R-CNN. Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2961–2969.	Pan, S. J., & Yang, Q. (2010). A survey on transfer learning. IEEE Transactions on Knowledge and Data Engineering, 22(10), 1345–1359.
Alias	pretrained instance segmentation, fine-tuned Mask R-CNN, transfer learning for panoptic segmentation, domain-adapted instance segmentation	pretrained object detector, fine-tuned object detection, TL-OD, domain-adapted object detection
Apparentées	4	3
Résumé	Transfer learning with instance segmentation reuses a backbone convolutional network pretrained on a large image corpus (typically ImageNet or COCO) as the feature extractor for an instance segmentation model such as Mask R-CNN, then fine-tunes the full pipeline on a smaller target dataset. This approach delivers state-of-the-art per-object mask accuracy with a fraction of the labeled data and compute that training from scratch would require.	Transfer learning with object detection starts from a deep neural network pretrained on a large image dataset — typically ImageNet for the backbone or COCO for the full detector — and adapts it to detect objects in a new domain. By reusing learned visual representations, it achieves strong detection accuracy with far fewer annotated images than training from scratch would require.