LangDA: Building Context-Awareness via Language for Domain Adaptive Semantic Segmentation
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866929761846034432 |
|---|---|
| author | Liu, Chang Balaji, Bavesh Hossain, Saad Thomas, C Lai, Kwei-Herng Vemulapalli, Raviteja Wong, Alexander Rambhatla, Sirisha |
| author_facet | Liu, Chang Balaji, Bavesh Hossain, Saad Thomas, C Lai, Kwei-Herng Vemulapalli, Raviteja Wong, Alexander Rambhatla, Sirisha |
| contents | Unsupervised domain adaptation for semantic segmentation (DASS) aims to transfer knowledge from a label-rich source domain to a target domain with no labels. Two key approaches in DASS are (1) vision-only approaches using masking or multi-resolution crops, and (2) language-based approaches that use generic class-wise prompts informed by target domain (e.g. "a {snowy} photo of a {class}"). However, the former is susceptible to noisy pseudo-labels that are biased to the source domain. The latter does not fully capture the intricate spatial relationships of objects -- key for dense prediction tasks. To this end, we propose LangDA. LangDA addresses these challenges by, first, learning contextual relationships between objects via VLM-generated scene descriptions (e.g. "a pedestrian is on the sidewalk, and the street is lined with buildings."). Second, LangDA aligns the entire image features with text representation of this context-aware scene caption and learns generalized representations via text. With this, LangDA sets the new state-of-the-art across three DASS benchmarks, outperforming existing methods by 2.6%, 1.4% and 3.9%. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_12780 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LangDA: Building Context-Awareness via Language for Domain Adaptive Semantic Segmentation Liu, Chang Balaji, Bavesh Hossain, Saad Thomas, C Lai, Kwei-Herng Vemulapalli, Raviteja Wong, Alexander Rambhatla, Sirisha Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Image and Video Processing 68Txx I.2.1 Unsupervised domain adaptation for semantic segmentation (DASS) aims to transfer knowledge from a label-rich source domain to a target domain with no labels. Two key approaches in DASS are (1) vision-only approaches using masking or multi-resolution crops, and (2) language-based approaches that use generic class-wise prompts informed by target domain (e.g. "a {snowy} photo of a {class}"). However, the former is susceptible to noisy pseudo-labels that are biased to the source domain. The latter does not fully capture the intricate spatial relationships of objects -- key for dense prediction tasks. To this end, we propose LangDA. LangDA addresses these challenges by, first, learning contextual relationships between objects via VLM-generated scene descriptions (e.g. "a pedestrian is on the sidewalk, and the street is lined with buildings."). Second, LangDA aligns the entire image features with text representation of this context-aware scene caption and learns generalized representations via text. With this, LangDA sets the new state-of-the-art across three DASS benchmarks, outperforming existing methods by 2.6%, 1.4% and 3.9%. |
| title | LangDA: Building Context-Awareness via Language for Domain Adaptive Semantic Segmentation |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Image and Video Processing 68Txx I.2.1 |
| url | https://arxiv.org/abs/2503.12780 |