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

Deep networks for stereo matching typically leverage 2D or 3D convolutionalencoder-decoder architectures to aggregate cost and regularize the cost volumefor accurate disparity estimation. Due to content-insensitive convolutions anddown-sampling and up-sampling operations, these cost aggregation mechanisms donot take full advantage of the information available in the images. Disparitymaps suffer from over-smoothing near occlusion boundaries, and erroneouspredictions in thin structures. In this paper, we show how deep adaptivefiltering and differentiable semi-global aggregation can be integrated inexisting 2D and 3D convolutional networks for end-to-end stereo matching,leading to improved accuracy. The improvements are due to utilizing RGBinformation from the images as a signal to dynamically guide the matchingprocess, in addition to being the signal we attempt to match across the images.We show extensive experimental results on the KITTI 2015 and Virtual KITTI 2datasets comparing four stereo networks (DispNetC, GCNet, PSMNet and GANet)after integrating four adaptive filters (segmentation-aware bilateralfiltering, dynamic filtering networks, pixel adaptive convolution andsemi-global aggregation) into their architectures. Our code is available athttps://github.com/ccj5351/DAFStereoNets.

Do End-to-end Stereo Algorithms Under-utilize Information?