Guided Distillation for Semi-Supervised Instance Segmentation - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2024

Guided Distillation for Semi-Supervised Instance Segmentation

Camille Couprie
  • Function : Author
  • PersonId : 764817
  • IdRef : 158438744
Jakob Verbeek

Abstract

Although instance segmentation methods have improved considerably, the dominant paradigm is to rely on fully-annotated training images, which are tedious to obtain. To alleviate this reliance, and boost results, semi-supervised approaches leverage unlabeled data as an additional training signal that limits overfitting to the labeled samples. In this context, we present novel design choices to significantly improve teacher-student distillation models. In particular, we (i) improve the distillation approach by introducing a novel “guided burn-in” stage, and (ii) evaluate different instance segmentation architectures, as well as backbone networks and pre-training strategies. Contrary to previous work which uses only supervised data for the burn-in period of the student model, we also use guidance of the teacher model to exploit unlabeled data in the burn-in period. Our improved distillation approach leads to substantial improvements over previous state-of-the-art results. For example, on the Cityscapes dataset we improve mask-AP from 23.7 to 33.9 when using labels for 10% of images, and on the COCO dataset we improve mask-AP from 18.3 to 34.1 when using labels for only 1% of the training data.
Fichier principal
Vignette du fichier
semi_supervised_instance_segmentation.pdf (4 Mo) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-04341872 , version 1 (13-12-2023)

Licence

Identifiers

Cite

Tariq Berrada, Camille Couprie, Karteek Alahari, Jakob Verbeek. Guided Distillation for Semi-Supervised Instance Segmentation. WACV 2024 - IEEE/CVF Winter Conference on Applications of Computer Vision, Jan 2024, Waikola, Hawaii, United States. pp.1-11. ⟨hal-04341872⟩
44 View
10 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More