Datasets:

MRAMG
/

MRAMG-Bench

ArXiv:

License:

Dataset card Files Files and versions

LbhYqh commited on Feb 5, 2025

Commit

ba8e263

·

verified ·

1 Parent(s): a7192d5

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -129,7 +129,7 @@ To evaluate the retrieval performance, we consider the following metrics:
 - **Visual Recall** measures the percentage of retrieved images relative to the total number of images in the ground truth.
   It is computed as:
-  ![123](https://huggingface.co/datasets/MRAMG/MRAMG-Bench/blob/main/cced8a3bb54ebd5906287f110a1e225.png)
   where "Retrieved Relevant Images" refers to the number of images retrieved that are present in the ground truth, and "Total Relevant Images in Ground Truth" refers to the total number of relevant images that should have been retrieved.

 - **Visual Recall** measures the percentage of retrieved images relative to the total number of images in the ground truth.
   It is computed as:
+  ![123](https://cdn-uploads.huggingface.co/production/uploads/67571051d39ac252085797ca/jQc-PXzlLsYix2_OhA3oP.png)
   where "Retrieved Relevant Images" refers to the number of images retrieved that are present in the ground truth, and "Total Relevant Images in Ground Truth" refers to the total number of relevant images that should have been retrieved.