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

Recent advancements in diffusion models and large language models (LLMs) havesignificantly propelled the field of AIGC. Text-to-Audio (TTA), a burgeoningAIGC application designed to generate audio from natural language prompts, isattracting increasing attention. However, existing TTA studies often strugglewith generation quality and text-audio alignment, especially for complextextual inputs. Drawing inspiration from state-of-the-art Text-to-Image (T2I)diffusion models, we introduce Auffusion, a TTA system adapting T2I modelframeworks to TTA task, by effectively leveraging their inherent generativestrengths and precise cross-modal alignment. Our objective and subjectiveevaluations demonstrate that Auffusion surpasses previous TTA approaches usinglimited data and computational resource. Furthermore, previous studies in T2Irecognizes the significant impact of encoder choice on cross-modal alignment,like fine-grained details and object bindings, while similar evaluation islacking in prior TTA works. Through comprehensive ablation studies andinnovative cross-attention map visualizations, we provide insightfulassessments of text-audio alignment in TTA. Our findings reveal Auffusion'ssuperior capability in generating audios that accurately match textualdescriptions, which further demonstrated in several related tasks, such asaudio style transfer, inpainting and other manipulations. Our implementationand demos are available at https://auffusion.github.io.

Auffusion: Leveraging the Power of Diffusion and Large Language Models for Text-to-Audio Generation