RepMet: Representative-based metric learning for classification and one-shot object detection

39 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Incremental Few-Shot Instance Segmentation (2021, methodology): “RepMet [30] trains a metric-learning sub-network to encode the support set, while Kang et al .”
Deep Learning for Generic Object Detection: A Survey (2018, unknown): “学習する少数ショット(few shot)検出の能力は非常に魅力的である [30, 61, 75, 129, 144, 228, 237].”
A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunities (2022, unknown): “One is to use generative models to caste prior knowledge [13], [14], [15], where bayesian programming learning [16] is the most representative framework [17] in this field.”
One-Shot Instance Segmentation (2018, background): “…closely related to our work is concurrent work based on Siamese networks for one-shot detection on an Omniglot-based dataset and for audio data [34] as well as work on ﬁne-grained bird clas-siﬁcation and localization in ImageNet images [77], which tend to have only one or few instances per image.”
Few-Shot Object Detection With Attention-RPN and Multi-Relation Detector (2019, methodology): “Which relation heads do we need? We follow the N way K-shot evaluation protocol proposed in RepMet [61] to evaluate our relation heads and other components.”