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

Many vision and language models suffer from poor visual grounding - oftenfalling back on easy-to-learn language priors rather than basing theirdecisions on visual concepts in the image. In this work, we propose a genericapproach called Human Importance-aware Network Tuning (HINT) that effectivelyleverages human demonstrations to improve visual grounding. HINT encouragesdeep networks to be sensitive to the same input regions as humans. Our approachoptimizes the alignment between human attention maps and gradient-based networkimportances - ensuring that models learn not just to look at but rather rely onvisual concepts that humans found relevant for a task when making predictions.We apply HINT to Visual Question Answering and Image Captioning tasks,outperforming top approaches on splits that penalize over-reliance on languagepriors (VQA-CP and robust captioning) using human attention demonstrations forjust 6% of the training data.

Taking a HINT: Leveraging Explanations to Make Vision and Language Models More Grounded