-1

I am looking for the State of the art deep learning model for facial expression classification in videos.

Given an Input video, I want the algorithm to detect the faces in them and classify the emotion associated to that face. The emotions that I am looking for are the standard 6 emotions + 1 neutral. The input video can be any self shot video and need not necessarily contain faces(As the algorithm will detect faces and then classify). Thus, for a single video, there can be multiple emotions associate

Kousik Krishnan
  • 118
  • 1
  • 6
  • What kind of classification are you looking at? What is the granularity that you want? – Rohit Shinde May 31 '17 at 20:52
  • Given an Input video, I want the algorithm to detect the faces in them and classify the emotion associated to that face. The emotions that I am looking for are the standard 6 emotions + 1 neutral. The input video can be any self shot video and need not necessarily contain faces(As the algorithm will detect faces and then classify). Thus, for a single video, there can be multiple emotions associated. – Kousik Krishnan Jun 01 '17 at 04:35

1 Answers1

0

Algorithm :

1) Train/get model for facial expression recognition for faces in images.

2) Extract individual frames from video

3) Detect the face in individual frames and pass the face to facial expression recognition for images and get the final output.

Kousik Krishnan
  • 118
  • 1
  • 6