Litcius/Paper detail

MimCo: Masked Image Modeling Pre-training with Contrastive Teacher

Qiang Zhou, Chaohui Yu, Hao Luo, Zhibin Wang, Hao Li

2022Proceedings of the 30th ACM International Conference on Multimedia16 citationsDOI

Abstract

Recent masked image modeling (MIM) has received much attention in self-supervised learning (SSL), which requires the target model to recover the masked part of the input image. Although MIM-based pre-training methods achieve new state-of-the-art performance when transferred to many downstream tasks, the visualizations show that the learned representations are less separable, especially compared to those based on contrastive learning pre-training. This inspires us to think whether the linear separability of MIM pre-trained representation can be further improved, thereby improving the pre-training performance. Since MIM and contrastive learning tend to utilize different data augmentations and training strategies, combining these two pretext tasks is not trivial. In this work, we propose a novel and flexible pre-training framework, named MimCo, which combines MIM and contrastive learning through two-stage pre-training. Specifically, MimCo takes a pre-trained contrastive learning model as the teacher model and is pre-trained with two types of learning targets: patch-level and image-level reconstruction losses.

Topics & Concepts

Computer scienceArtificial intelligenceImage (mathematics)Training (meteorology)PretextRepresentation (politics)Training setNatural language processingMachine learningPattern recognition (psychology)LawPolitical sciencePhysicsMeteorologyPoliticsDomain Adaptation and Few-Shot LearningGenerative Adversarial Networks and Image SynthesisAdvanced Neural Network Applications