diff --git a/IFRT/README.md b/IFRT/README.md
index 40162f3..5eb5a6b 100644
--- a/IFRT/README.md
+++ b/IFRT/README.md
@@ -1,7 +1,7 @@
 # InsightFace Recognition Test (IFRT)
 **IFRT** is a globalised fair benchmark for face recognition algorithms. IFRT evaluates the algorithm performance on worldwide web pictures which contain various sex, age and race groups, but no identification photos.
 
-**IFRT** testset consists of non-celebrities so we can ensure that it contains very few overlap with public available face recognition training set, such as MS1M and CASIA as they mostly come from online celebrities' photo. As the result, we can evaluate the `fair` performance for different algorithms.
+**IFRT** testset consists of non-celebrities so we can ensure that it has very few overlap with public available face recognition training set, such as MS1M and CASIA as they mostly collected from online celebrities' photo. As the result, we can evaluate the *FAIR* performance for different algorithms.
 
 Similar to [FRVT](https://www.nist.gov/programs-projects/face-recognition-vendor-test-frvt), we encourage participants to prepare a black-box feature extractor. To simplify this process, users can just replace their trained ArcFace model(with or without encryption) in our simple open-sourced pre-packaged software.