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

Recognition and reasoning are two pillars of visual understanding. However,these tasks have an imbalance in focus; whereas recent advances in neuralnetworks have shown strong empirical performance in visual recognition, therehas been comparably much less success in solving visual reasoning. Intuitively,unifying these two tasks under a singular framework is desirable, as they aremutually dependent and beneficial. Motivated by the recent success ofmulti-task transformers for visual recognition and language understanding, wepropose a unified neural architecture for visual recognition and reasoning witha generic interface (e.g., tokens) for both. Our framework enables theprincipled investigation of how different visual recognition tasks, datasets,and inductive biases can help enable spatiotemporal reasoning capabilities.Noticeably, we find that object detection, which requires spatial localizationof individual objects, is the most beneficial recognition task for reasoning.We further demonstrate via probing that implicit object-centric representationsemerge automatically inside our framework. Intriguingly, we discover thatcertain architectural choices such as the backbone model of the visual encoderhave a significant impact on visual reasoning, but little on object detection.Given the results of our experiments, we believe that visual reasoning shouldbe considered as a first-class citizen alongside visual recognition, as theyare strongly correlated but benefit from potentially different design choices.

Towards A Unified Neural Architecture for Visual Recognition and Reasoning