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

The success of supervised learning requires large-scale ground truth labelswhich are very expensive, time-consuming, or may need special skills toannotate. To address this issue, many self- or un-supervised methods aredeveloped. Unlike most existing self-supervised methods to learn only 2D imagefeatures or only 3D point cloud features, this paper presents a novel andeffective self-supervised learning approach to jointly learn both 2D imagefeatures and 3D point cloud features by exploiting cross-modality andcross-view correspondences without using any human annotated labels.Specifically, 2D image features of rendered images from different views areextracted by a 2D convolutional neural network, and 3D point cloud features areextracted by a graph convolution neural network. Two types of features are fedinto a two-layer fully connected neural network to estimate the cross-modalitycorrespondence. The three networks are jointly trained (i.e. cross-modality) byverifying whether two sampled data of different modalities belong to the sameobject, meanwhile, the 2D convolutional neural network is additionallyoptimized through minimizing intra-object distance while maximizinginter-object distance of rendered images in different views (i.e. cross-view).The effectiveness of the learned 2D and 3D features is evaluated bytransferring them on five different tasks including multi-view 2D shaperecognition, 3D shape recognition, multi-view 2D shape retrieval, 3D shaperetrieval, and 3D part-segmentation. Extensive evaluations on all the fivedifferent tasks across different datasets demonstrate strong generalization andeffectiveness of the learned 2D and 3D features by the proposed self-supervisedmethod.

Self-supervised Feature Learning by Cross-modality and Cross-view Correspondences