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

Source-free cross-modal knowledge transfer is a crucial yet challenging task,which aims to transfer knowledge from one source modality (e.g., RGB) to thetarget modality (e.g., depth or infrared) with no access to the task-relevant(TR) source data due to memory and privacy concerns. A recent attempt leveragesthe paired task-irrelevant (TI) data and directly matches the features fromthem to eliminate the modality gap. However, it ignores a pivotal clue that thepaired TI data could be utilized to effectively estimate the source datadistribution and better facilitate knowledge transfer to the target modality.To this end, we propose a novel yet concise framework to unlock the potentialof paired TI data for enhancing source-free cross-modal knowledge transfer. Ourwork is buttressed by two key technical components. Firstly, to better estimatethe source data distribution, we introduce a Task-irrelevant data-GuidedModality Bridging (TGMB) module. It translates the target modality data (e.g.,infrared) into the source-like RGB images based on paired TI data and theguidance of the available source model to alleviate two key gaps: 1)inter-modality gap between the paired TI data; 2) intra-modality gap between TIand TR target data. We then propose a Task-irrelevant data-Guided KnowledgeTransfer (TGKT) module that transfers knowledge from the source model to thetarget model by leveraging the paired TI data. Notably, due to theunavailability of labels for the TR target data and its less reliableprediction from the source model, our TGKT model incorporates a self-supervisedpseudo-labeling approach to enable the target model to learn from itspredictions. Extensive experiments show that our method achievesstate-of-the-art performance on three datasets (RGB-to-depth andRGB-to-infrared).

Source-Free Cross-Modal Knowledge Transfer by Unleashing the Potential of Task-Irrelevant Data