Real-time face detection and emotion/gender classification using fer2013/IMDB datasets & Head-pose detection to measure roll, pitch & yaw using a keras CNN model and openCV3

• IMDB gender classification test accuracy: 96%.
• fer2013 emotion classification test accuracy: 66%.

Instructions for Emotion Recognition + Gender Recognition

Run real-time emotion demo:

python3 video_emotion_color_demo.py

To train previous/new models for emotion classification:

• Download the fer2013.tar.gz file from here

• Move the downloaded file to the datasets directory inside this repository.

• Untar the file:

tar -xzf fer2013.tar

• Run the train_emotion_classification.py file

python3 train_emotion_classifier.py

To train previous/new models for gender classification:

• Download the imdb_crop.tar file from here (It's the 7GB button with the title Download faces only).

• Move the downloaded file to the datasets directory inside this repository.

• Untar the file:

tar -xfv imdb_crop.tar

• Run the train_gender_classification.py file

python3 train_gender_classifier.py

Instructions for Head Orientation Detection

Research Paper

Patacchiola, M., & Cangelosi, A. (2017). Head pose estimation in the wild using Convolutional Neural Networks and adaptive gradient methods. Pattern Recognition, http://dx.doi.org/10.1016/j.patcog.2017.06.009.

Run real-time head-pose detection demo:

python3 head_pose_estimation_webcam.py