Add paper link and metadata for ESPnet (#1)

- Add paper link and metadata for ESPnet (944e30c33e21b8d95bf514a8ecf45dd3d066e05b)


Co-authored-by: Niels Rogge <nielsr@users.noreply.huggingface.co>

README.md

@@ -1,13 +1,25 @@
 ---
+datasets:
+- as2m
+license: cc-by-4.0
+library_name: espnet
+pipeline_tag: audio-classification
 tags:
 - espnet
 - audio
 - self-supervised-learning
-datasets:
-- as2m
-license: cc-by-4.0
 ---
 
+# OpenBEATs-ICME
+
+This repository contains the audio encoder model presented in the paper [The CMU-AIST submission for the ICME 2025 Audio Encoder Challenge](https://huggingface.co/papers/2601.16273).
+
+## Model Description
+The system is built on BEATs, a masked speech token prediction-based audio encoder. This version scales the architecture up to 300 million parameters and was pre-trained using 74,000 hours of audio data derived from various speech, music, and sound corpora.
+
+- **Code:** [ESPnet GitHub](https://github.com/espnet/espnet/)
+- **Paper:** [The CMU-AIST submission for the ICME 2025 Audio Encoder Challenge](https://huggingface.co/papers/2601.16273)
+
 ## ESPnet2 SSL model
 
 ### `shikhar7ssu/OpenBEATs-ICME`
@@ -1241,12 +1253,6 @@ distributed: true
   doi={10.21437/Interspeech.2018-1456},
   url={http://dx.doi.org/10.21437/Interspeech.2018-1456}
 }
-
-
-
-
-
-
 ```
 
 or arXiv:
@@ -1260,4 +1266,4 @@ or arXiv:
   archivePrefix={arXiv},
   primaryClass={cs.CL}
 }
-```
+```