MixMatch Domain Adaptaion: Prize-winning solution for both tracks of\n VisDA 2019 challenge

2019/10/09 by Д. И. Рухович, Rukhovich, Danila, Danil Galeev +1 · 4 citations
Computer Science · Biochemistry, Genetics and Molecular Biology · #Domain Adaptation and Few-Shot Learning #Cancer-related molecular mechanisms research #Speech Recognition and Synthesis

paper · pdf · doi:10.48550/arxiv.1910.03903

Abstract

We present a domain adaptation (DA) system that can be used in multi-source\nand semi-supervised settings. Using the proposed method we achieved 2nd place\non multi-source track and 3rd place on semi-supervised track of the VisDA 2019\nchallenge (http://ai.bu.edu/visda-2019/). The source code of the method is\navailable at https://github.com/filaPro/visda2019.\n

Cited by

• Semi-supervised Domain Adaptation for Semantic Segmentation
• Source Data-absent Unsupervised Domain Adaptation through Hypothesis Transfer and Labeling Transfer
• Domain Adaptation with Auxiliary Target Domain-Oriented Classifier
• Semi-supervised Domain Adaptation based on Dual-level Domain Mixing for Semantic Segmentation

Related

• Multi-Source Domain Adaptation and Semi-Supervised Domain Adaptation with Focus on Visual Domain Adaptation Challenge 2019
• Pre-Training Transformers for Domain Adaptation
• CoNMix for Source-free Single and Multi-target Domain Adaptation
• AdaEmbed: Semi-supervised Domain Adaptation in the Embedding Space
• 3rd Place Solution for VisDA 2021 Challenge -- Universally Domain Adaptive Image Recognition