Rewrite Caption Semantics: Bridging Semantic Gaps for Language-Supervised Semantic Segmentation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Xing, Yun, Kang, Jian, Xiao, Aoran, Nie, Jiahao, Shao, Ling, Lu, Shijian
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866929197144866816
author Xing, Yun
Kang, Jian
Xiao, Aoran
Nie, Jiahao
Shao, Ling
Lu, Shijian
author_facet Xing, Yun
Kang, Jian
Xiao, Aoran
Nie, Jiahao
Shao, Ling
Lu, Shijian
contents Vision-Language Pre-training has demonstrated its remarkable zero-shot recognition ability and potential to learn generalizable visual representations from language supervision. Taking a step ahead, language-supervised semantic segmentation enables spatial localization of textual inputs by learning pixel grouping solely from image-text pairs. Nevertheless, the state-of-the-art suffers from clear semantic gaps between visual and textual modality: plenty of visual concepts appeared in images are missing in their paired captions. Such semantic misalignment circulates in pre-training, leading to inferior zero-shot performance in dense predictions due to insufficient visual concepts captured in textual representations. To close such semantic gap, we propose Concept Curation (CoCu), a pipeline that leverages CLIP to compensate for the missing semantics. For each image-text pair, we establish a concept archive that maintains potential visually-matched concepts with our proposed vision-driven expansion and text-to-vision-guided ranking. Relevant concepts can thus be identified via cluster-guided sampling and fed into pre-training, thereby bridging the gap between visual and textual semantics. Extensive experiments over a broad suite of 8 segmentation benchmarks show that CoCu achieves superb zero-shot transfer performance and greatly boosts language-supervised segmentation baseline by a large margin, suggesting the value of bridging semantic gap in pre-training data.
format Preprint
id arxiv_https___arxiv_org_abs_2309_13505
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Rewrite Caption Semantics: Bridging Semantic Gaps for Language-Supervised Semantic Segmentation
Xing, Yun
Kang, Jian
Xiao, Aoran
Nie, Jiahao
Shao, Ling
Lu, Shijian
Computer Vision and Pattern Recognition
Vision-Language Pre-training has demonstrated its remarkable zero-shot recognition ability and potential to learn generalizable visual representations from language supervision. Taking a step ahead, language-supervised semantic segmentation enables spatial localization of textual inputs by learning pixel grouping solely from image-text pairs. Nevertheless, the state-of-the-art suffers from clear semantic gaps between visual and textual modality: plenty of visual concepts appeared in images are missing in their paired captions. Such semantic misalignment circulates in pre-training, leading to inferior zero-shot performance in dense predictions due to insufficient visual concepts captured in textual representations. To close such semantic gap, we propose Concept Curation (CoCu), a pipeline that leverages CLIP to compensate for the missing semantics. For each image-text pair, we establish a concept archive that maintains potential visually-matched concepts with our proposed vision-driven expansion and text-to-vision-guided ranking. Relevant concepts can thus be identified via cluster-guided sampling and fed into pre-training, thereby bridging the gap between visual and textual semantics. Extensive experiments over a broad suite of 8 segmentation benchmarks show that CoCu achieves superb zero-shot transfer performance and greatly boosts language-supervised segmentation baseline by a large margin, suggesting the value of bridging semantic gap in pre-training data.
title Rewrite Caption Semantics: Bridging Semantic Gaps for Language-Supervised Semantic Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2309.13505