Filter Images First, Generate Instructions Later: Pre-Instruction Data Selection for Visual Instruction Tuning
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_ | 1866915228712697856 |
|---|---|
| author | Safaei, Bardia Siddiqui, Faizan Xu, Jiacong Patel, Vishal M. Lo, Shao-Yuan |
| author_facet | Safaei, Bardia Siddiqui, Faizan Xu, Jiacong Patel, Vishal M. Lo, Shao-Yuan |
| contents | Visual instruction tuning (VIT) for large vision-language models (LVLMs) requires training on expansive datasets of image-instruction pairs, which can be costly. Recent efforts in VIT data selection aim to select a small subset of high-quality image-instruction pairs, reducing VIT runtime while maintaining performance comparable to full-scale training. However, a major challenge often overlooked is that generating instructions from unlabeled images for VIT is highly expensive. Most existing VIT datasets rely heavily on human annotations or paid services like the GPT API, which limits users with constrained resources from creating VIT datasets for custom applications. To address this, we introduce Pre-Instruction Data Selection (PreSel), a more practical data selection paradigm that directly selects the most beneficial unlabeled images and generates instructions only for the selected images. PreSel first estimates the relative importance of each vision task within VIT datasets to derive task-wise sampling budgets. It then clusters image features within each task, selecting the most representative images with the budget. This approach reduces computational overhead for both instruction generation during VIT data formation and LVLM fine-tuning. By generating instructions for only 15% of the images, PreSel achieves performance comparable to full-data VIT on the LLaVA-1.5 and Vision-Flan datasets. The link to our project page: https://bardisafa.github.io/PreSel |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_07591 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Filter Images First, Generate Instructions Later: Pre-Instruction Data Selection for Visual Instruction Tuning Safaei, Bardia Siddiqui, Faizan Xu, Jiacong Patel, Vishal M. Lo, Shao-Yuan Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Visual instruction tuning (VIT) for large vision-language models (LVLMs) requires training on expansive datasets of image-instruction pairs, which can be costly. Recent efforts in VIT data selection aim to select a small subset of high-quality image-instruction pairs, reducing VIT runtime while maintaining performance comparable to full-scale training. However, a major challenge often overlooked is that generating instructions from unlabeled images for VIT is highly expensive. Most existing VIT datasets rely heavily on human annotations or paid services like the GPT API, which limits users with constrained resources from creating VIT datasets for custom applications. To address this, we introduce Pre-Instruction Data Selection (PreSel), a more practical data selection paradigm that directly selects the most beneficial unlabeled images and generates instructions only for the selected images. PreSel first estimates the relative importance of each vision task within VIT datasets to derive task-wise sampling budgets. It then clusters image features within each task, selecting the most representative images with the budget. This approach reduces computational overhead for both instruction generation during VIT data formation and LVLM fine-tuning. By generating instructions for only 15% of the images, PreSel achieves performance comparable to full-data VIT on the LLaVA-1.5 and Vision-Flan datasets. The link to our project page: https://bardisafa.github.io/PreSel |
| title | Filter Images First, Generate Instructions Later: Pre-Instruction Data Selection for Visual Instruction Tuning |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2503.07591 |