InSight is able to track tiny movements of facial muscles in a single-user’s face and translate them into universal facial expressions like happiness, surprise, sadness, anger and more. Additionally, the SDK can uniquely track eye movements with high accuracy and precision in order to understand users’ attention patterns and which visual stimuli grabs their interest. All this by using a simple camera.
InSight is an extremely flexible and cross-platform software toolkit which can work on the most popular desktop and mobile platforms (Windows, Mac, Linux, iOS, Android).
InSight accurately measures relevant facial data like facial expressions, age, gender, head pose and eye gaze location.
You can measure the emotional reactions and engagement level of your users while they are watching your videos, advertisement or website content, and classify this information per age or gender group.
Offer customized experiences to your users, create new ways of interacting with content and devices, by using emotion recognition and eye tracking technologies.
You are now able to monitor which parts of a game induce frustration or satisfaction, or to even command a virtual character according to their emotions or gaze.
The extensive range of features InSight measures, can provide your company with powerful face analytics and a deep knowledge about a single individual.
The collected face analytics, enables you to understand what your users like and what they ignore, therefore, being able to make more informed and cost-effective Marketing decisions.
Measure 7 universal emotions with high accuracy: happy, sad, surprised, scared, disgusted, puzzled, neutral
Age and Gender Classification
Extract relevant demographics such as age and gender.
Real time Results
Process the webcam stream and get real-time information on emotional behavior and eye gaze attention of your users.
Define where your users focus their visual attention with high precision.
Head Pose Estimation
Retrieve 6-DOF (x, y, z, yaw, pitch, roll) head pose estimation .
Retrieve real-time information from individuals using your favorite platform: Windows, Mac, Linux, iOS, Android.
Features and Accuracy
Feature Accuracy Facial Expressions Average: 93.2 %
Neutral : 88.2 %
Happy : 95.2 %
Surprised : 100.0 %
Puzzled : 98.3 %
Disgusted : 81.1 %
Afraid : 94.2 %
Sad : 95.6 %
Eye Gaze Tracking ~ 2.1° Age Estimation +/- 7 years Gender Estimation 86 % Head Pose Estimation pitch 5.2° (+/- 4.6°)
yaw 6.1° (+/- 5.79°)
roll 3.00° (+/- 2.82°)
Market ResearchGet reliable insights on consumers emotional behavior and interest towards content, products and brands, without the need to inquire them. Learn More.
Creative IndustriesOffer interactive gaming experiences and art installations which reflects your audience/consumers emotional changes and interest. Learn More.
ClinicalUnderstand changes on emotional behavior and attention level, for an early diagnosis of neurological diseases. Learn More.
User ExperienceOffer improved user experience and website usability, by understanding which visual stimuli grabbed your users’ attention and drove their emotional reactions. Learn More.
What does SDK means?SDK stands for Software Development Kit. This means that it is a technology toolbox (library) which can be used by programmers to integrate our technologies into new or existing applications.
What is the SDK designed for?The SDK is designed for Human-Computer Interaction and Human Behavior Analysis. We consider the SDK to be an enabling toolbox for many applications; it is up to you to find great ideas on how you can use these technologies.
Can I track multiple people at the same time?No, the SDK is targeted to a single person scenarios. If there are multiple people in the image, the SDK only considers the face closer to the camera. If this feature is a necessity, please consider using CrowdSight SDK.
What programming languages are supported?The SDK is a C++ library, this can be wrapped to support more languages, but the wrapping is not directly supported by Sightcorp.
What is the minimum system requirement?The SDK is able to work offline on basically every system. For real-time performance, we suggest to use at least an Intel core 2 Duo @ 2.5 Ghz. A full HD web camera (eg: Logitech HD pro series) is required for enhanced tracking accuracy. An internet connection is required for activation and license validation, depending on your license type.
What operating systems are supported?Windows 7, Linux (Ubuntu), Mac (Snow Leopard or better), Android and iOS
Does the SDK work on people with beards, sunglasses or spectacles?Yes, the SDK is robust against these features. However, sunglasses might affect the emotional classification (that is why poker player wear sunglasses!). Also, sunglasses and reflection on spectacles does not yield reliable eye tracking results.
What kinds of illumination condition are required?The SDK works with a various set of conditions. The optimal condition is with frontal, homogeneous natural illumination. Avoid backlighting and make sure the camera is not over-saturated. The SDK also works in the dark. As a rule of thumb, if the face features can be clearly seen from the input image, the SDK should be able to interpret them.
Do you have a scientific papers describing the inner working and functionality of the SDK?The various technologies included in the SDK are on the works for as much as 15 years by scientists at the University of Amsterdam. Although some of the papers are available in the literature, updates and improvements achieved over the last years are not released to the public.
Click here for full screen access to the documentation
For a proper use of the InSight SDK, the following minimal technical capabilities are required:
- Intel Core 2 Duo 2.0GHZ or better
- 2GB RAM
- 100MB free hard disk space
- 640×480 resolution webcam (Suggested: Logitech HD Pro Webcam C910)
- User distance of approximately 60 cm from the camera