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

Multi-task learning (MTL) seeks to learn a single model to accomplishmultiple tasks by leveraging shared information among the tasks. Existing MTLmodels, however, have been known to suffer from negative interference amongtasks. Efforts to mitigate task interference have focused on eitherloss/gradient balancing or implicit parameter partitioning with partialoverlaps among the tasks. In this paper, we propose ETR-NLP to mitigate taskinterference through a synergistic combination of non-learnable primitives(NLPs) and explicit task routing (ETR). Our key idea is to employ non-learnableprimitives to extract a diverse set of task-agnostic features and recombinethem into a shared branch common to all tasks and explicit task-specificbranches reserved for each task. The non-learnable primitives and the explicitdecoupling of learnable parameters into shared and task-specific ones affordthe flexibility needed for minimizing task interference. We evaluate theefficacy of ETR-NLP networks for both image-level classification andpixel-level dense prediction MTL problems. Experimental results indicate thatETR-NLP significantly outperforms state-of-the-art baselines with fewerlearnable parameters and similar FLOPs across all datasets. Code is availableat this \href{https://github.com/zhichao-lu/etr-nlp-mtl}.

Mitigating Task Interference in Multi-Task Learning via Explicit Task Routing with Non-Learnable Primitives