fixing README typos

Readme.md

@@ -23,11 +23,11 @@ Although having initially trained with 10 different identities, we can only publ
 
 If you want to create your own dataset, it works as follows:  
 
-- Run `get_data.py`. It requires the path of a .csv file with the following information: the youtube url `Channel-URL`, the name of the youtuber `Name` and the gender `Gender`. The output dataset will consist on a folder for each of the youtubers specified in the .csv file as well as two folders (video/audio) in each of them containing the full video/audio downloaded.  
+- Run `get_data.py`. It requires the path of a .csv file with the following information: the youtube url `Channel-URL`, the name of the youtuber `Name` and the gender `Gender`. The output dataset will consist on a folder for each different youtuber specified in the .csv file as well as two folders (video/audio) in each of them containing the full video/audio downloaded.  
 
 - Run `preprocess_audio_frames.py` for audio preprocessing purposes.
 
-- Remove any file whose name does not finish like `*preprocessing_wav` in each audio folder.  
+- Remove any file whose name does not finish like `preprocessing_wav` in each audio folder.  
 
 - Run `face_detector.py` specifying the dataset path, a confidence threshold needed to compute the bounding boxes and the path of the haarcascade pretrained classifier (included in this repo in `/assets`)