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

Deep motion forecasting models have achieved great success when trained on amassive amount of data. Yet, they often perform poorly when training data islimited. To address this challenge, we propose a transfer learning approach forefficiently adapting pre-trained forecasting models to new domains, such asunseen agent types and scene contexts. Unlike the conventional fine-tuningapproach that updates the whole encoder, our main idea is to reduce the amountof tunable parameters that can precisely account for the target domain-specificmotion style. To this end, we introduce two components that exploit our priorknowledge of motion style shifts: (i) a low-rank motion style adapter thatprojects and adjusts the style features at a low-dimensional bottleneck; and(ii) a modular adapter strategy that disentangles the features of scene contextand motion history to facilitate a fine-grained choice of adaptation layers.Through extensive experimentation, we show that our proposed adapter design,coined MoSA, outperforms prior methods on several forecasting benchmarks.

Motion Style Transfer: Modular Low-Rank Adaptation for Deep Motion Forecasting