diff --git a/recognition/arcface_torch/README.md b/recognition/arcface_torch/README.md
index e6f92c5ba..ea27365c1 100644
--- a/recognition/arcface_torch/README.md
+++ b/recognition/arcface_torch/README.md
@@ -164,12 +164,4 @@ More details see
   booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
   year={2022}
 }
-@inproceedings{an2020partical_fc,
-  title={Partial FC: Training 10 Million Identities on a Single Machine},
-  author={An, Xiang and Zhu, Xuhan and Xiao, Yang and Wu, Lan and Zhang, Ming and Gao, Yuan and Qin, Bin and
-  Zhang, Debing and Fu Ying},
-  booktitle={Proceedings of International Conference on Computer Vision Workshop},
-  pages={1445-1449},
-  year={2020}
-}
 ```
diff --git a/recognition/arcface_torch/partial_fc.py b/recognition/arcface_torch/partial_fc.py
index 4e74279bf..93ead8b4a 100644
--- a/recognition/arcface_torch/partial_fc.py
+++ b/recognition/arcface_torch/partial_fc.py
@@ -8,7 +8,7 @@
 
 class PartialFC(torch.nn.Module):
     """
-    https://arxiv.org/abs/2010.05222
+    https://arxiv.org/abs/2203.15565
     A distributed sparsely updating variant of the FC layer, named Partial FC (PFC).
 
     When sample rate less than 1, in each iteration, positive class centers and a random subset of