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

Activity classification has observed great success recently. The performanceon small dataset is almost saturated and people are moving towards largerdatasets. What leads to the performance gain on the model and what the modelhas learnt? In this paper we propose identity preserve transform (IPT) to studythis problem. IPT manipulates the nuisance factors (background, viewpoint,etc.) of the data while keeping those factors related to the task (humanmotion) unchanged. To our surprise, we found popular models are using highlycorrelated information (background, object) to achieve high classificationaccuracy, rather than using the essential information (human motion). This canexplain why an activity classification model usually fails to generalize todatasets it is not trained on. We implement IPT in two forms, i.e. image-spacetransform and 3D transform, using synthetic images. The tool will be madeopen-source to help study model and dataset design.

Identity Preserve Transform: Understand What Activity Classification Models Have Learnt