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

Deep neural networks (DNNs) have achieved great success in the area ofcomputer vision. The disparity estimation problem tends to be addressed by DNNswhich achieve much better prediction accuracy than traditional hand-craftedfeature-based methods. However, the existing DNNs hardly serve both efficientcomputation and rich expression capability, which makes them difficult fordeployment in real-time and high-quality applications, especially on mobiledevices. To this end, we propose an efficient, accurate, and configurable deepnetwork for disparity estimation named FADNet++. Leveraging several liberalnetwork design and training techniques, FADNet++ can boost its accuracy with afast model inference speed for real-time applications. Besides, it enablesusers to easily configure different sizes of models for balancing accuracy andinference efficiency. We conduct extensive experiments to demonstrate theeffectiveness of FADNet++ on both synthetic and realistic datasets among sixGPU devices varying from server to mobile platforms. Experimental results showthat FADNet++ and its variants achieve state-of-the-art prediction accuracy,and run at a significant order of magnitude faster speed than existing 3Dmodels. With the constraint of running at above 15 frames per second (FPS) on amobile GPU, FADNet++ achieves a new state-of-the-art result for the SceneFlowdataset.

FADNet++: Real-Time and Accurate Disparity Estimation with Configurable Networks