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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Compared with ample visual-text pre-training research, few works exploreaudio-text pre-training, mostly due to the lack of sufficient parallelaudio-text data. Most existing methods incorporate the visual modality as apivot for audio-text pre-training, which inevitably induces data noise. In thispaper, we propose BLAT: Bootstrapping Language-Audio pre-training based onTag-guided synthetic data. We utilize audio captioning to generate textdirectly from audio, without the aid of the visual modality so that potentialnoise from modality mismatch is eliminated. Furthermore, we propose captiongeneration under the guidance of AudioSet tags, leading to more accuratecaptions. With the above two improvements, we curate high-quality, large-scaleparallel audio-text data, based on which we perform audio-text pre-training.Evaluation on a series of downstream tasks indicates that BLAT achieves SOTAzero-shot classification performance on most datasets and significantperformance improvement when fine-tuned on downstream tasks, suggesting theeffectiveness of our synthetic data.

BLAT: Bootstrapping Language-Audio Pre-training based on AudioSet Tag-guided Synthetic Data