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

Conventional supervised learning methods typically assume i.i.d samples andare found to be sensitive to out-of-distribution (OOD) data. We proposeGenerative Causal Representation Learning (GCRL) which leverages causality tofacilitate knowledge transfer under distribution shifts. While we evaluate theeffectiveness of our proposed method in human trajectory prediction models,GCRL can be applied to other domains as well. First, we propose a novel causalmodel that explains the generative factors in motion forecasting datasets usingfeatures that are common across all environments and with features that arespecific to each environment. Selection variables are used to determine whichparts of the model can be directly transferred to a new environment withoutfine-tuning. Second, we propose an end-to-end variational learning paradigm tolearn the causal mechanisms that generate observations from features. GCRL issupported by strong theoretical results that imply identifiability of thecausal model under certain assumptions. Experimental results on synthetic andreal-world motion forecasting datasets show the robustness and effectiveness ofour proposed method for knowledge transfer under zero-shot and low-shotsettings by substantially outperforming the prior motion forecasting models onout-of-distribution prediction. Our code is available athttps://github.com/sshirahmad/GCRL.

Generative Causal Representation Learning for Out-of-Distribution Motion Forecasting