Sampled Dataset for: A Unified Vision-Language Model for Cross-Product Defect Detection in Glove Manufacturing

This repository contains the sampled dataset supporting the research article: "A Unified Vision-Language Model for Cross-Product Defect Detection in Glove Manufacturing", which has been accepted for publication in PLOS ONE. The dataset includes sampled images and label files required to replicate the training phases described in the study. The data is organized into two compressed archives: qwen2.5-sft-dataset-1k.zip: This file contains the dataset used for the Supervised Fine-Tuning (SFT) stage.qwen2.5-rft-dataset-8k.zip: This file contains the dataset used for the Reinforcement Fine-Tuning (RFT) stage.Associated Publication: Zhao, Y., Tian, L., & Wang, Y. "A Unified Vision-Language Model for Cross-Product Defect Detection in Glove Manufacturing." PLOS ONE (Accepted/In Press), 2025.