Repository for final implementation of all methods accepted in Blended joint Attention repo. Report an issue before creating a pull request.
This repository deals with work done by The Distibuted Red Hen Lab towards classification of different instances of blended classic joint attention in various form of print, audio and video media. For more information visit : https://sites.google.com/site/distributedlittleredhen/home/the-cognitive-core-research-topics-in-red-hen/the-barnyard/blended-classic-joint-attention
Face detection : Detection of number of human faces, possible extensions to their position and orientation Emotion recognition : Recognising different emotions (sad, happy, surprised, neutral etc.) using a CNN classifier Gaze direction recognition : Calculating angle of ones gaze using initial pupil detection and terminal points of eyes. Age detection : Categorising a person's age via facial features (outputs a range of possible age values) Facial Landmark detection : Detecting major facial landmarks, which is useful for Gaze direction and Emotion recognition. Blended CLassic Joint attention : Detectiong instances of BCJA from instances without BCJA Reaction Shots : Analyse reaction shots (of surprise, awe etc.) Gesture Recognition : Recognising multimodal gestures Required Packages:
- Python 2.7.x
- Numpy
- Bob
- Matplotlib
- OpenCV (One must check compatibility with python and OS)
- DLib
- pympi-ling
- Dr.Mark Turner
- Dr.Francis Steen
- Soumitra Agarwal
- Debayan Das