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

Scene graph generation (SGG) is to detect object pairs with their relationsin an image. Existing SGG approaches often use multi-stage pipelines todecompose this task into object detection, relation graph construction, anddense or dense-to-sparse relation prediction. Instead, from a perspective onSGG as a direct set prediction, this paper presents a simple, sparse, andunified framework, termed as Structured Sparse R-CNN. The key to our method isa set of learnable triplet queries and a structured triplet detector whichcould be jointly optimized from the training set in an end-to-end manner.Specifically, the triplet queries encode the general prior for object pairswith their relations, and provide an initial guess of scene graphs forsubsequent refinement. The triplet detector presents a cascaded architecture toprogressively refine the detected scene graphs with the customized dynamicheads. In addition, to relieve the training difficulty of our method, wepropose a relaxed and enhanced training strategy based on knowledgedistillation from a Siamese Sparse R-CNN. We perform experiments on severaldatasets: Visual Genome and Open Images V4/V6, and the results demonstrate thatour method achieves the state-of-the-art performance. In addition, we alsoperform in-depth ablation studies to provide insights on our structuredmodeling in triplet detector design and training strategies. The code andmodels are made available at https://github.com/MCG-NJU/Structured-Sparse-RCNN.

Structured Sparse R-CNN for Direct Scene Graph Generation