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

We consider the problem of next frame prediction from video input. Arecurrent convolutional neural network is trained to predict depth frommonocular video input, which, along with the current video image and the cameratrajectory, can then be used to compute the next frame. Unlike prior next-frameprediction approaches, we take advantage of the scene geometry and use thepredicted depth for generating the next frame prediction. Our approach canproduce rich next frame predictions which include depth information attached toeach pixel. Another novel aspect of our approach is that it predicts depth froma sequence of images (e.g. in a video), rather than from a single still image.We evaluate the proposed approach on the KITTI dataset, a standard dataset forbenchmarking tasks relevant to autonomous driving. The proposed method producesresults which are visually and numerically superior to existing methods thatdirectly predict the next frame. We show that the accuracy of depth predictionimproves as more prior frames are considered.

Geometry-Based Next Frame Prediction from Monocular Video