Instruction-Grounded Visual Projectors for Continual Learning of Generative Vision-Language Models
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866918109563060224 |
|---|---|
| author | Jin, Hyundong Chang, Hyung Jin Kim, Eunwoo |
| author_facet | Jin, Hyundong Chang, Hyung Jin Kim, Eunwoo |
| contents | Continual learning enables pre-trained generative vision-language models (VLMs) to incorporate knowledge from new tasks without retraining data from previous ones. Recent methods update a visual projector to translate visual information for new tasks, connecting pre-trained vision encoders with large language models. However, such adjustments may cause the models to prioritize visual inputs over language instructions, particularly learning tasks with repetitive types of textual instructions. To address the neglect of language instructions, we propose a novel framework that grounds the translation of visual information on instructions for language models. We introduce a mixture of visual projectors, each serving as a specialized visual-to-language translation expert based on the given instruction context to adapt to new tasks. To avoid using experts for irrelevant instruction contexts, we propose an expert recommendation strategy that reuses experts for tasks similar to those previously learned. Additionally, we introduce expert pruning to alleviate interference from the use of experts that cumulatively activated in previous tasks. Extensive experiments on diverse vision-language tasks demonstrate that our method outperforms existing continual learning approaches by generating instruction-following responses. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_00260 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Instruction-Grounded Visual Projectors for Continual Learning of Generative Vision-Language Models Jin, Hyundong Chang, Hyung Jin Kim, Eunwoo Computer Vision and Pattern Recognition Multimedia Continual learning enables pre-trained generative vision-language models (VLMs) to incorporate knowledge from new tasks without retraining data from previous ones. Recent methods update a visual projector to translate visual information for new tasks, connecting pre-trained vision encoders with large language models. However, such adjustments may cause the models to prioritize visual inputs over language instructions, particularly learning tasks with repetitive types of textual instructions. To address the neglect of language instructions, we propose a novel framework that grounds the translation of visual information on instructions for language models. We introduce a mixture of visual projectors, each serving as a specialized visual-to-language translation expert based on the given instruction context to adapt to new tasks. To avoid using experts for irrelevant instruction contexts, we propose an expert recommendation strategy that reuses experts for tasks similar to those previously learned. Additionally, we introduce expert pruning to alleviate interference from the use of experts that cumulatively activated in previous tasks. Extensive experiments on diverse vision-language tasks demonstrate that our method outperforms existing continual learning approaches by generating instruction-following responses. |
| title | Instruction-Grounded Visual Projectors for Continual Learning of Generative Vision-Language Models |
| topic | Computer Vision and Pattern Recognition Multimedia |
| url | https://arxiv.org/abs/2508.00260 |