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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Recently end-to-end unsupervised deep learning methods have achieved aneffect beyond geometric methods for visual depth and ego-motion estimationtasks. These data-based learning methods perform more robustly and accuratelyin some of the challenging scenes. The encoder-decoder network has been widelyused in the depth estimation and the RCNN has brought significant improvementsin the ego-motion estimation. Furthermore, the latest use of GenerativeAdversarial Nets(GANs) in depth and ego-motion estimation has demonstrated thatthe estimation could be further improved by generating pictures in the gamelearning process. This paper proposes a novel unsupervised network system forvisual depth and ego-motion estimation: Stacked Generative AdversarialNetwork(SGANVO). It consists of a stack of GAN layers, of which the lowestlayer estimates the depth and ego-motion while the higher layers estimate thespatial features. It can also capture the temporal dynamic due to the use of arecurrent representation across the layers. See Fig.1 for details. We selectthe most commonly used KITTI [1] data set for evaluation. The evaluationresults show that our proposed method can produce better or comparable resultsin depth and ego-motion estimation.

SGANVO: Unsupervised Deep Visual Odometry and Depth Estimation with Stacked Generative Adversarial Networks