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

Reasoning about visual relationships is central to how humans interpret thevisual world. This task remains challenging for current deep learningalgorithms since it requires addressing three key technical problems jointly:1) identifying object entities and their properties, 2) inferring semanticrelations between pairs of entities, and 3) generalizing to novelobject-relation combinations, i.e., systematic generalization. In this work, weuse vision transformers (ViTs) as our base model for visual reasoning and makebetter use of concepts defined as object entities and their relations toimprove the reasoning ability of ViTs. Specifically, we introduce a novelconcept-feature dictionary to allow flexible image feature retrieval attraining time with concept keys. This dictionary enables two new concept-guidedauxiliary tasks: 1) a global task for promoting relational reasoning, and 2) alocal task for facilitating semantic object-centric correspondence learning. Toexamine the systematic generalization of visual reasoning models, we introducesystematic splits for the standard HICO and GQA benchmarks. We show theresulting model, Concept-guided Vision Transformer (or RelViT for short)significantly outperforms prior approaches on HICO and GQA by 16% and 13% inthe original split, and by 43% and 18% in the systematic split. Our ablationanalyses also reveal our model's compatibility with multiple ViT variants androbustness to hyper-parameters.

RelViT: Concept-guided Vision Transformer for Visual Relational Reasoning