@inproceedings{preeti-etal-2026-semantically,
    title = "Semantically Aware Optimal Transport for Dense Label Transfer",
    author = "Preeti  and
      Ravish, Kiran  and
      Kushwaha, Ankita  and
      Kumar, Pawan",
    editor = "Yan, Qianqi  and
      Montariol, Syrielle  and
      Fan, Yue  and
      Gu, Jing  and
      Pan, Jiayi  and
      Li, Manling  and
      Kordjamshidi, Parisa  and
      Suhr, Alane  and
      Wang, Xin Eric",
    booktitle = "Proceedings of the 4th Workshop on Advances in Language and Vision Research ({ALVR})",
    month = jul,
    year = "2026",
    address = "San Diego, California, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl-workshops/2026.alvr-main.3/",
    pages = "18--45",
    ISBN = "979-8-89176-398-2",
    abstract = "Vision foundation models produce features that generalize across visual domains without fine-tuning, yet naively transferring labels through these feature spaces fails under large distribution shifts.We propose SAOT (**S**emantically **A**ware **O**ptimal **T**ransport), which learns a transport cost within a fused unbalanced optimal transport formulation for dense label transfer from frozen vision transformer features to new domains.SAOT combines a learnable appearance metric with semantic class-prototype priors, unbalanced transport for partial matching under distribution shift, and a block-sparse solver for tractable inference.We pair this with a two-stage decoder: an MLP trained on SAOT pseudo-labels, then refined via EMA-teacher self-training with class-balanced sampling.On GTA5$\to$Cityscapes with frozen DINOv2 ViT-L/14 features, SAOT+Decoder reaches 25.7{\%} mIoU, a **3.8$\times$** improvement over nearest-neighbor transfer (6.7{\%}), without any backbone adaptation.Per-class results show large gains on spatially coherent classes (road 90.3{\%}, car 76.2{\%}, building 71.5{\%}), demonstrating that learned semantic transport costs capture domain-invariant structure even under severe synthetic-to-real shifts. On VOC train$\to$val with frozen ViT-B/16 features, the full pipeline reaches 47.5{\%} mIoU, indicating that the approach extends beyond synthetic-to-real adaptation."
}