The document discusses Affectiva's Emotion SDK for Linux. It provides an overview of the SDK's capabilities including supported emotion metrics like smile, surprise, sadness, detectors for different media sources, data structures for frames and faces, and callbacks. It also outlines the steps to track faces in an application using the SDK and how to access emotion metrics. Key requirements like supported Linux distributions, programming languages, and hardware are also mentioned. The presentation encourages developers to visit Affectiva's developer portal and GitHub for documentation, sample code, and support.
3. @affectiva #emodev16
Software and Hardware Requirements
Supported Linux Distros:
Ubuntu 14.04 LTS
CentOS 7
Programming Language:
C++
Processor Architecture:
x86_64, ARM (coming soon!)
Hardware Requirements: (recommended)
Processor, 2GHz
RAM 1GHz
Compiler Requirements:
GLIBC 3.4, GLIBCXX 3.4.9, GCC v4.8
4. @affectiva #emodev16
SDK Detectors
affdex::PhotoDetector affdex::FrameDetector affdex::CameraDetector affdex::VideoDetector
Source
Detector
For each of the different sources, a child of the class affdex::Detector can be used to consume
pixels and produce emotion metrics.
5. @affectiva #emodev16
SDK Data Structures
The affdex::Frame data structure contains the
raw pixels of the image in BGR Format.
6. @affectiva #emodev16
The affdex::Face data structure contains the
results for the analysis of the different
expressions, emotions and locations of the face
for an individual frame.
SDK Data Structures
7. @affectiva #emodev16
SDK Callbacks
The SDK uses callback functions to report
events.
• affdex::ImageListener: receive image and
emotion analysis
• affdex::FaceListener: receive events for
every face found and lost.
9. @affectiva #emodev16
How do Track Faces in my app ?
(1) Create an instance of the detector matching
your data source. Specify the number of faces to
track, or just track one.
10. @affectiva #emodev16
(2) Specify the location of the Affdex Emotion
SDK data folder. The location of the classifier
files to use.
How do Track Faces in my app ?
12. @affectiva #emodev16
(4) Implement affdex::FaceListener callback to
receive notification when a face is found or lost.
How do Track Faces in my app ?
16. @affectiva #emodev16
How to get started ?
Developer portal:
Documentation and how to guides
to show how to use the SDKs
Sample code and utilities on GitHub
Ask questions through stack exchange
tag: affdex-sdk
/Affectiva
Visit: http://developer.affectiva.com