I'm a beginner to the machine learning and data science. Somehow clear with the theoretical notions, but as a beginner i'm finding it a bit harder to implement. I've small query 'Whether the schema i'm trying to develop will work or not?'
I'm building a model to predict the emotional status using facial feature extraction. I've a model for that in the picture.
Actually I've got a set of images associated with the particular emotions. Basically i'm trying to perform the classification on the top. I'm detecting the edges of the eyes, mouth and performing regression over the pixel position for such edges.
My confusion: Whether the regression over the position would be efficient or not. If not what are the best ways to perform such task. I'm not concerned about neural nets or any other. I'm just trying to put it simple as possible.
Note: Please do help me to improve my questioning, so i can get a bit closer to what i'm asking. As a beginner i mightn't be able to express my query properly.