English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Unsupervised cross-domain action recognition aims at adapting the modeltrained on an existing labeled source domain to a new unlabeled target domain.Most existing methods solve the task by directly aligning the featuredistributions of source and target domains. However, this would cause negativetransfer during domain adaptation due to some negative training samples in bothdomains. In the source domain, some training samples are of low-relevance totarget domain due to the difference in viewpoints, action styles, etc. In thetarget domain, there are some ambiguous training samples that can be easilyclassified as another type of action under the case of source domain. Theproblem of negative transfer has been explored in cross-domain objectdetection, while it remains under-explored in cross-domain action recognition.Therefore, we propose a Multi-modal Instance Refinement (MMIR) method toalleviate the negative transfer based on reinforcement learning. Specifically,a reinforcement learning agent is trained in both domains for every modality torefine the training data by selecting out negative samples from each domain.Our method finally outperforms several other state-of-the-art baselines incross-domain action recognition on the benchmark EPIC-Kitchens dataset, whichdemonstrates the advantage of MMIR in reducing negative transfer.

Multi-modal Instance Refinement for Cross-domain Action Recognition