{"id":"arxiv/2604.26370","url":"https://arxiv.org/abs/2604.26370","tags":["cs.CV","cs.LG","math.AT"],"score":0,"title":"Topology-Aware Representation Alignment for Semi-Supervised Vision-Language Learning","domain":"cs.CV","source":"arxiv","authors":["Junwon You","Mihyun Jang","Sangwoo Mo","Jae-Hun Jung"],"reviews":0,"subject":"cs.CV","summary":"Vision-language models have shown strong performance, but they often generalize poorly to specialized domains. While semi-supervised vision-language learning mitigates this limitation by leveraging a small set of labeled image-text pairs together with abundant unlabeled images, existing methods remain fundamentally pairwise and fail to model the global structure of multimodal representation manifolds. Most existing topology-based alignment methods rely on persistence diagram matching, which neither guarantees geometric alignment nor uses the image-text pairing information central to…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-04-29","reactions":{}}