Datasets:

Modalities:
Image
Size:
< 1K
Libraries:
Datasets
License:
Weiyun1025 commited on
Commit
254def4
1 Parent(s): 2159f33

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -10,7 +10,7 @@ The evaluation is formulated as single-choice questions.
10
  The benchmark consists of four splits:
11
  **Existence**, **Subject**, **Predicate**, and **Object**.
12
 
13
- The **Existence** split evaluates the object recognition ability while the remaining splits are designed to evaluate the capability of relation comprehension, focusing on probing each of the elements in the relation triplets `(subject, predicate, object)` of the scene graph separately.
14
  Some data examples are shown below.
15
 
16
  <img width="800" alt="image" src="https://cdn-uploads.huggingface.co/production/uploads/619507e7b74b6c591f794340/_NKaowl2OUBAjck1XCAPm.jpeg">
 
10
  The benchmark consists of four splits:
11
  **Existence**, **Subject**, **Predicate**, and **Object**.
12
 
13
+ The **Existence** split evaluates the object recognition ability while the remaining splits are designed to evaluate the capability of relation comprehension, focusing on probing each of the elements in the relation triplets `(subject, predicate, object)` separately.
14
  Some data examples are shown below.
15
 
16
  <img width="800" alt="image" src="https://cdn-uploads.huggingface.co/production/uploads/619507e7b74b6c591f794340/_NKaowl2OUBAjck1XCAPm.jpeg">