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

Many-to-many multimodal summarization (M$^3$S) task aims to generatesummaries in any language with document inputs in any language and thecorresponding image sequence, which essentially comprises multimodalmonolingual summarization (MMS) and multimodal cross-lingual summarization(MXLS) tasks. Although much work has been devoted to either MMS or MXLS and hasobtained increasing attention in recent years, little research pays attentionto the M$^3$S task. Besides, existing studies mainly focus on 1) utilizing MMSto enhance MXLS via knowledge distillation without considering the performanceof MMS or 2) improving MMS models by filtering summary-unrelated visualfeatures with implicit learning or explicitly complex training objectives. Inthis paper, we first introduce a general and practical task, i.e., M$^3$S.Further, we propose a dual knowledge distillation and target-oriented visionmodeling framework for the M$^3$S task. Specifically, the dual knowledgedistillation method guarantees that the knowledge of MMS and MXLS can betransferred to each other and thus mutually prompt both of them. To offertarget-oriented visual features, a simple yet effective target-orientedcontrastive objective is designed and responsible for discarding needlessvisual information. Extensive experiments on the many-to-many setting show theeffectiveness of the proposed approach. Additionally, we will contribute amany-to-many multimodal summarization (M$^3$Sum) dataset.

D$^2$TV: Dual Knowledge Distillation and Target-oriented Vision Modeling for Many-to-Many Multimodal Summarization