CA-LoRA: Concept-Aware LoRA for Domain-Aligned Segmentation Dataset Generation
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_ | 1866918407850426368 |
|---|---|
| author | Park, Minho Park, Sunghyun Lee, Jungsoo Park, Hyojin Hwang, Kyuwoong Porikli, Fatih Choo, Jaegul Choi, Sungha |
| author_facet | Park, Minho Park, Sunghyun Lee, Jungsoo Park, Hyojin Hwang, Kyuwoong Porikli, Fatih Choo, Jaegul Choi, Sungha |
| contents | This paper addresses the challenge of data scarcity in semantic segmentation by generating datasets through text-to-image (T2I) generation models, reducing image acquisition and labeling costs. Segmentation dataset generation faces two key challenges: 1) aligning generated samples with the target domain and 2) producing informative samples beyond the training data. Fine-tuning T2I models can help generate samples aligned with the target domain. However, it often overfits and memorizes training data, limiting their ability to generate diverse and well-aligned samples. To overcome these issues, we propose Concept-Aware LoRA (CA-LoRA), a novel fine-tuning approach that selectively identifies and updates only the weights associated with necessary concepts (e.g., style or viewpoint) for domain alignment while preserving the pretrained knowledge of the T2I model to produce informative samples. We demonstrate its effectiveness in generating datasets for urban-scene segmentation, outperforming baseline and state-of-the-art methods in in-domain (few-shot and fully-supervised) settings, as well as in domain generalization tasks, especially under challenging conditions such as adverse weather and varying illumination, further highlighting its superiority. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_22172 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CA-LoRA: Concept-Aware LoRA for Domain-Aligned Segmentation Dataset Generation Park, Minho Park, Sunghyun Lee, Jungsoo Park, Hyojin Hwang, Kyuwoong Porikli, Fatih Choo, Jaegul Choi, Sungha Computer Vision and Pattern Recognition This paper addresses the challenge of data scarcity in semantic segmentation by generating datasets through text-to-image (T2I) generation models, reducing image acquisition and labeling costs. Segmentation dataset generation faces two key challenges: 1) aligning generated samples with the target domain and 2) producing informative samples beyond the training data. Fine-tuning T2I models can help generate samples aligned with the target domain. However, it often overfits and memorizes training data, limiting their ability to generate diverse and well-aligned samples. To overcome these issues, we propose Concept-Aware LoRA (CA-LoRA), a novel fine-tuning approach that selectively identifies and updates only the weights associated with necessary concepts (e.g., style or viewpoint) for domain alignment while preserving the pretrained knowledge of the T2I model to produce informative samples. We demonstrate its effectiveness in generating datasets for urban-scene segmentation, outperforming baseline and state-of-the-art methods in in-domain (few-shot and fully-supervised) settings, as well as in domain generalization tasks, especially under challenging conditions such as adverse weather and varying illumination, further highlighting its superiority. |
| title | CA-LoRA: Concept-Aware LoRA for Domain-Aligned Segmentation Dataset Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.22172 |