Nano1337/FPS-scores-positive-VQAnotFPS

该数据集包含100个样本，主要用于图像与文本描述的相关性分析。每个样本包含用户ID（uid）、图像（image）、文本描述（caption）、FP分数（FPscore）、VQA分数（VQAscore）等字段。FPscore和VQAscore分别表示图像与文本描述的匹配度，以及通过视觉问答模型评估的分数。此外，数据集中还包含了FPscore和VQAscore的百分位数（FPscore_percentile、VQAscore_percentile）以及两者的百分位数差异（percentile_difference）。数据集仅包含一个训练集（train），占用4056221字节，下载大小为4023484字节。

This dataset contains 100 samples and is primarily used for analyzing the correlation between images and text descriptions. Each sample includes fields such as user ID (uid), image, caption, FPscore, VQAscore, etc. FPscore and VQAscore represent the matching degree between the image and text description, and the score evaluated by a visual question answering model, respectively. Additionally, the dataset includes percentiles for FPscore and VQAscore (FPscore_percentile, VQAscore_percentile) and the difference between these percentiles (percentile_difference). The dataset contains only a training set (train), occupying 4056221 bytes, with a download size of 4023484 bytes.