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

Temporal action detection aims to recognize the action category and determinethe starting and ending time of each action instance in untrimmed videos. Themixed methods have achieved remarkable performance by simply merginganchor-based and anchor-free approaches. However, there are still two crucialissues in the mixed framework: (1) Brute-force merging and handcrafted anchorsdesign affect the performance and practical application of the mixed methods.(2) A large number of false positives in action category predictions furtherimpact the detection performance. In this paper, we propose a novel BoundaryDiscretization and Reliable Classification Network (BDRC-Net) that addressesthe above issues by introducing boundary discretization and reliableclassification modules. Specifically, the boundary discretization module (BDM)elegantly merges anchor-based and anchor-free approaches in the form ofboundary discretization, avoiding the handcrafted anchors design required bytraditional mixed methods. Furthermore, the reliable classification module(RCM) predicts reliable action categories to reduce false positives in actioncategory predictions. Extensive experiments conducted on different benchmarksdemonstrate that our proposed method achieves favorable performance comparedwith the state-of-the-art. For example, BDRC-Net hits an average mAP of 68.6%on THUMOS'14, outperforming the previous best by 1.5%. The code will bereleased at https://github.com/zhenyingfang/BDRC-Net.

Boundary Discretization and Reliable Classification Network for Temporal Action Detection