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

Underwater acoustic target recognition (UATR) and localization (UATL) playimportant roles in marine exploration. The highly noisy acoustic signal andtime-frequency interference among various sources pose big challenges to thistask. To tackle these issues, we propose a multimodal approach to extract andfuse audio-visual-textual information to recognize and localize underwatertargets through the designed Symbiotic Transformer (Symb-Trans) and Multi-ViewRegression (MVR) method. The multimodal data were first preprocessed by acustom-designed HetNorm module to normalize the multi-source data in a commonfeature space. The Symb-Trans module embeds audiovisual features by co-trainingthe preprocessed multimodal features through parallel branches and a contentencoder with cross-attention. The audiovisual features are then used forunderwater target recognition. Meanwhile, the text embedding combined with theaudiovisual features is fed to an MVR module to predict the localization of theunderwater targets through multi-view clustering and multiple regression. Sinceno off-the-shell multimodal dataset is available for UATR and UATL, we combinedmultiple public datasets, consisting of acoustic, and/or visual, and/ortextural data, to obtain audio-visual-textual triplets for model training andvalidation. Experiments show that our model outperforms comparative methods in91.7% (11 out of 12 metrics) and 100% (4 metrics) of the quantitative metricsfor the recognition and localization tasks, respectively. In a case study, wedemonstrate the advantages of multi-view models in establishing samplediscriminability through visualization methods. For UATL, the proposed MVRmethod produces the relation graphs, which allow predictions based on recordsof underwater targets with similar conditions.

More Perspectives Mean Better: Underwater Target Recognition and  Localization with Multimodal Data via Symbiotic Transformer and Multiview  Regression