LocCa: Visual Pretraining with Location-aware Captioners

TL;DR

LocCa introduces a location-aware captioning pretraining method that enhances localization capabilities in image understanding models, outperforming standard captioners on localization tasks while maintaining overall performance.

Contribution

The paper presents a novel pretraining approach incorporating location-aware captioners, leveraging multitask encoder-decoder architecture for improved localization in image captioning.

Findings

Outperforms standard captioners on localization tasks

Maintains comparable performance on holistic image captioning tasks

Demonstrates effectiveness of location-aware pretraining

Abstract

Image captioning has been shown as an effective pretraining method similar to contrastive pretraining. However, the incorporation of location-aware information into visual pretraining remains an area with limited research. In this paper, we propose a simple visual pretraining method with location-aware captioners (LocCa). LocCa uses a simple image captioner task interface, to teach a model to read out rich information, i.e. bounding box coordinates, and captions, conditioned on the image pixel input. Thanks to the multitask capabilities of an encoder-decoder architecture, we show that an image captioner can easily handle multiple tasks during pretraining. Our experiments demonstrate that LocCa outperforms standard captioners significantly on localization downstream tasks while maintaining comparable performance on holistic tasks.