Update README.md

Codes/README.md

@@ -22,11 +22,23 @@
 # To Run the unimodal Vision Based models
 
 6.Vision+lstm_foldWise.py   
-7. 3DCNN_withFolds.py
+7.3DCNN_withFolds.py
 
 # To Run the Multimodal Model
        
 9. MultiModalFusionModelfoldWise.py
 
 # To extract all the video frames.
 frameExtract.py
+
+# Extraction of transcript
+
+The 'all__video_vosk_audioMap.p' has to be generated using the Vosk speech recognition toolkit(https://alphacephei.com/vosk/). The format of the file is in JSON format like the below:
+
+{
+  "video_name1": "transcript1",
+  "video_name2": "transcript2",
+  ...
+  "video_name3": "transcript3"
+}
+