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

In the era of end-to-end deep learning, many advances in computer vision aredriven by large amounts of labeled data. In the optical flow setting, however,obtaining dense per-pixel ground truth for real scenes is difficult and thussuch data is rare. Therefore, recent end-to-end convolutional networks foroptical flow rely on synthetic datasets for supervision, but the domainmismatch between training and test scenarios continues to be a challenge.Inspired by classical energy-based optical flow methods, we design anunsupervised loss based on occlusion-aware bidirectional flow estimation andthe robust census transform to circumvent the need for ground truth flow. Onthe KITTI benchmarks, our unsupervised approach outperforms previousunsupervised deep networks by a large margin, and is even more accurate thansimilar supervised methods trained on synthetic datasets alone. By optionallyfine-tuning on the KITTI training data, our method achieves competitive opticalflow accuracy on the KITTI 2012 and 2015 benchmarks, thus in addition enablinggeneric pre-training of supervised networks for datasets with limited amountsof ground truth.

UnFlow: Unsupervised Learning of Optical Flow with a Bidirectional Census Loss