UP-DP: Unsupervised Prompt Learning for Data Pre-Selection with Vision-Language Models
–Neural Information Processing Systems
In this study, we investigate the task of data pre-selection, which aims to select instances for labeling from an unlabeled dataset through a single pass, thereby optimizing performance for undefined downstream tasks with a limited annotation budget. Previous approaches to data pre-selection relied solely on visual features extracted from foundation models, such as CLIP and BLIP-2, but largely ignored the powerfulness of text features. In this work, we argue that, with proper design, the joint feature space of both vision and text can yield a better representation for data pre-selection. To this end, we introduce UP-DP, a simple yet effective unsupervised prompt learning approach that adapts vision-language models, like BLIP-2, for data pre-selection. Specifically, with the BLIP-2 parameters frozen, we train text prompts to extract the joint features with improved representation, ensuring a diverse cluster structure that covers the entire dataset. We extensively compare our method with the state-of-the-art using seven benchmark datasets in different settings, achieving up to a performance gain of 20%. Interestingly, the prompts learned from one dataset demonstrate significant generalizability and can be applied directly to enhance the feature extraction of BLIP-2 from other datasets. To the best of our knowledge, UP-DP is the first work to incorporate unsupervised prompt learning in a vision-language model for data pre-selection.
Neural Information Processing Systems
Apr-24-2026, 09:50:41 GMT
- Genre:
- Research Report > New Finding (0.34)
- Technology:
- Information Technology > Artificial Intelligence
- Vision (1.00)
- Natural Language (1.00)
- Machine Learning
- Statistical Learning (1.00)
- Inductive Learning (0.94)
- Neural Networks > Deep Learning (0.46)
- Information Technology > Artificial Intelligence