Data Distillation: Towards Omni-Supervised Learning
Dec 20, 2017
Data Distillation: Towards Omni-Supervised Learning
-
Radosavovic, Ilija;
-
Dollár, Piotr;
-
Girshick, Ross;
-
Gkioxari, Georgia;
-
He, Kaiming;
Abstract: We investigate omni-supervised learning, a special regime of semi-supervised learning in which the learner exploits all available labeled data plus internet-scale sources of unlabeled data. Omni-supervised learning is lower-bounded by performance on existing labeled datasets, offering the potential to surpass state-of-the-art fully supervised methods. To exploit the omni-supervised setting, we propose data distillation, a method that ensembles predictions from multiple transformations of unlabeled data, using a single model, to automatically generate new training annotations. We argue that visual recognition models have recently become accurate enough that it is now possible to apply classic ideas about self-training to challenging real-world data. Our experimental results show that in the cases of human keypoint detection and general object detection, state-of-the-art models trained with data distillation surpass the performance of using labeled data from the COCO dataset alone.
https://arxiv.org/abs/1712.04440v1
← Back to all articles Quick Navigation: Next:[ j ] – Prev:[ k ] – List:[ l ]