Update README.md

README.md

@@ -71,6 +71,16 @@ Automatic speech recognition: word error rate, character error rate, phone error
 ![distance_wer_plot](./png2pdf-3-1-1.png)
 ![shout_wer_plot](./png2pdf-4-1-1.png)
 
+Speech emotion recognition: F1 Score
+
+| Model                  | UA ↑   | WA ↑   |
+|------------------------|--------|--------|
+| SpeechBrain Wav2Vec2   | 20.7%  | 20.8%  |
+| DAWN-hidden-SVM        | **32.1%** | **32.2%** |
+| Wav2Small-VAD-SVM*     | 23.3%  | 22.3%  |
+
+*Teacher model
+SVM indicates an SVM on the hidden layers or VAD output, see paper for details
 
 ## Metadata Details