Abstract:
When we talk about the human emotion the human face act as a very important in
terms of finding an individual’s mood or emotion. There are various emotions such
as happy, sad, angry, etc which can be identified with help of facial expressions. till
now if the user wants to make the playlist they have to go through the list of the music
then select the songs based on their emotions but it takes consumes more time and
it becomes a very tedious and upheld task for the user. Previously many algorithms
have been proposed for generating the songs automatically. but the conventional
algorithms which are in use are required various external hardware or sensors like
electroencephalogram for capturing and identifying the human emotion via human
brain it makes the complete process very slow and less accurate. existing systems are
not user-friendly they have the complex architecture however This proposed system
based on extracted facial expression is user-friendly any user can use it anywhere
any time. also proposed system eliminate the task of manually creating the playlists
of songs based on the emotions it automatically generates the different playlist
It saves much more time and efforts of users who are music lover. Thus the proposed
system (Emo-music) aims to minimize the computational time as compared
to existing algorithms for getting the results it also reduces the overall cost of the
designed system, thereby given features will automatically increase the overall accuracy
of the proposed system. The proposed system (Emo-music) tested on both
utilize-dependent and utilize-independent datasets. Visages are captured utilizing an
inbuilt camera. The precision or Accuracy of the emotion detection algorithm utilized
in the system is around 80-95%. Thus, it yields better precision compared to
the algorithms utilized in the literature survey.
Keywords: Android,Human Face, Emotional Features, PlayList ,User Independent
Dataset,User Dependent Dataset, Emotion detection, Inbuilt Camera,Emotion
Recognition , Face Recognition,Songsextraction,web scrapping,youtube-dl.
