Update README.md

README.md

@@ -22,7 +22,25 @@ This cleaned dataset forms a core part of the training data for our OWSM v4 mode
 - **Model Demo:** [Gradio](https://huggingface.co/spaces/espnet/OWSM_V4_Demo)
 
 
-## OWSM v4 Models
+## OWSM v4 Results
+
+Please refer to our paper for comprehensive evaluations. Below are some notable results.
+
+### Language Identification
+
+![Language identification accuracy](lid-result.png)
+
+
+### English ASR
+
+![English ASR WER vs inference speed](en-asr.png)
+
+### Multilingual ASR
+
+![Multilingual ASR WER](fleurs.png)
+
+
+## Pre-trained Models
 
 ### Encoder-decoder OWSM