ChimeraLoRA: Multi-Head LoRA-Guided Synthetic Datasets
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866908874851745792 |
|---|---|
| author | Kim, Hoyoung Jang, Minwoo Koo, Jabin Yun, Sangdoo Ok, Jungseul |
| author_facet | Kim, Hoyoung Jang, Minwoo Koo, Jabin Yun, Sangdoo Ok, Jungseul |
| contents | Beyond general recognition tasks, specialized domains and fine-grained settings often encounter data scarcity, especially for tail classes. To obtain less biased and more reliable models under such scarcity, practitioners leverage diffusion models to supplement underrepresented regions of real data. Specifically, recent studies fine-tune pretrained diffusion models with LoRA on few-shot real sets to synthesize additional images. While an image-wise LoRA trained on a single image captures fine-grained details yet offers limited diversity, a class-wise LoRA trained over all shots produces diverse images as it encodes class priors yet tends to overlook fine details. To combine both benefits, we separate the adapter into a class-shared LoRA~$A$ for class priors and per-image LoRAs~$\mathcal{B}$ for image-specific characteristics. To expose coherent class semantics in the shared LoRA~$A$, we propose a semantic boosting by preserving class bounding boxes during training. For generation, we compose $A$ with a mixture of $\mathcal{B}$ using coefficients drawn from a Dirichlet distribution. Across diverse datasets, our synthesized images are both diverse and detail-rich while closely aligning with the few-shot real distribution, yielding robust gains in downstream classification accuracy. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_19708 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | ChimeraLoRA: Multi-Head LoRA-Guided Synthetic Datasets Kim, Hoyoung Jang, Minwoo Koo, Jabin Yun, Sangdoo Ok, Jungseul Computer Vision and Pattern Recognition Beyond general recognition tasks, specialized domains and fine-grained settings often encounter data scarcity, especially for tail classes. To obtain less biased and more reliable models under such scarcity, practitioners leverage diffusion models to supplement underrepresented regions of real data. Specifically, recent studies fine-tune pretrained diffusion models with LoRA on few-shot real sets to synthesize additional images. While an image-wise LoRA trained on a single image captures fine-grained details yet offers limited diversity, a class-wise LoRA trained over all shots produces diverse images as it encodes class priors yet tends to overlook fine details. To combine both benefits, we separate the adapter into a class-shared LoRA~$A$ for class priors and per-image LoRAs~$\mathcal{B}$ for image-specific characteristics. To expose coherent class semantics in the shared LoRA~$A$, we propose a semantic boosting by preserving class bounding boxes during training. For generation, we compose $A$ with a mixture of $\mathcal{B}$ using coefficients drawn from a Dirichlet distribution. Across diverse datasets, our synthesized images are both diverse and detail-rich while closely aligning with the few-shot real distribution, yielding robust gains in downstream classification accuracy. |
| title | ChimeraLoRA: Multi-Head LoRA-Guided Synthetic Datasets |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.19708 |