Implementing Adaptations for Vision AutoRegressive Model
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916866721579008 |
|---|---|
| author | Shaikh, Kaif Boenisch, Franziska Dziedzic, Adam |
| author_facet | Shaikh, Kaif Boenisch, Franziska Dziedzic, Adam |
| contents | Vision AutoRegressive model (VAR) was recently introduced as an alternative to Diffusion Models (DMs) in image generation domain. In this work we focus on its adaptations, which aim to fine-tune pre-trained models to perform specific downstream tasks, like medical data generation. While for DMs there exist many techniques, adaptations for VAR remain underexplored. Similarly, differentially private (DP) adaptations-ones that aim to preserve privacy of the adaptation data-have been extensively studied for DMs, while VAR lacks such solutions. In our work, we implement and benchmark many strategies for VAR, and compare them to state-of-the-art DM adaptation strategies. We observe that VAR outperforms DMs for non-DP adaptations, however, the performance of DP suffers, which necessitates further research in private adaptations for VAR. Code is available at https://github.com/sprintml/finetuning_var_dp. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_11441 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Implementing Adaptations for Vision AutoRegressive Model Shaikh, Kaif Boenisch, Franziska Dziedzic, Adam Computer Vision and Pattern Recognition Machine Learning I.2.6; I.5.1; I.4.8; I.2.10 Vision AutoRegressive model (VAR) was recently introduced as an alternative to Diffusion Models (DMs) in image generation domain. In this work we focus on its adaptations, which aim to fine-tune pre-trained models to perform specific downstream tasks, like medical data generation. While for DMs there exist many techniques, adaptations for VAR remain underexplored. Similarly, differentially private (DP) adaptations-ones that aim to preserve privacy of the adaptation data-have been extensively studied for DMs, while VAR lacks such solutions. In our work, we implement and benchmark many strategies for VAR, and compare them to state-of-the-art DM adaptation strategies. We observe that VAR outperforms DMs for non-DP adaptations, however, the performance of DP suffers, which necessitates further research in private adaptations for VAR. Code is available at https://github.com/sprintml/finetuning_var_dp. |
| title | Implementing Adaptations for Vision AutoRegressive Model |
| topic | Computer Vision and Pattern Recognition Machine Learning I.2.6; I.5.1; I.4.8; I.2.10 |
| url | https://arxiv.org/abs/2507.11441 |