Unity AI

Face SDK for Unity 3D.

MoodMe Face Recognition SDKs use AI neural networks adapted to Unity Barracuda.

Product availability: Windows 10, Android, iOS, MacOS & Web.

MoodMe Unity SDKs empower Unity developers to create applications that detect emotions, recognize human faces’ characteristics, and apply Augmented Reality filters.

We adapt our product to all new Unity updates.

MoodMe Emotion SDK

MoodMe Unity Emotion detection SDK of MoodMe works as a standard Unity 3D plugin and runs on all platforms supported by Unity.

The SDK takes the video footage from the standard Webcam Texture Unity class. Furthermore, it can be fed by any kind of image or video converted in any supported format (RGB, RGBA, BGRA, YUV, YUY2).

Its output for each frame (image) are 7 emotions values (happy surprise afraid angry disgust sad & neutral) on a 0-100 range.

Read more about our Emotion detection Neural Network here.

MoodMe Face AR SDK specifications

MoodMe Unity Face Recognition plugin works as a standard Unity 3D plugin with specific extensions for Windows 64 bit, OSX, Android, and iOS. The Unity Face Recognition plugin takes the video footage from the standard Webcam Texture Unity class. Furthermore, it can be fed by any kind of image or video converted in any supported format (RGB, RGBA, BGRA, YUV, YUY2).

Unity Face Recognition SDK plugin is a thread-safe multi-thread plugin that can be scaled to use every available CPU core. It runs at 60 FPS on mid-end systems and up to 85 FPS on high-end systems.

CUDA and OpenCL are supported.

The Unity Face Recognition plugin makes the 66/68 2D/3D landmarks data available to Unity. It also supplies the head transformation matrix. Animation drivers and retargeting available.

Unity Face Recognition Emotion detection SDK enables Game Developers to create scenarios where players’ moods can be detected and applied to their in-game characters in real-time. Furthermore, the mood of the player is detected and made available as 7 different values (for each base emotion: happy, surprise, angry, sad, afraid, disgust, neutral).

Consequently, the gaming experience can be enriched with adaptive scenarios tailored to how the player is feeling. While the games did a lot of guesswork, gaming companies using Unity3D and MoodMe Unity Face Recognition family of SDKs can now create games that adapt to the gamers’ moods.

For privacy reasons YouTube needs your permission to be loaded.
I Accept

Emotions AI on Unity Store

MoodMe Emotion Detection free trial is available directly from the Unity Asset Store.

Contact [email protected] for 7-Emotions, AR, Gender, Age & Ethnicity SDKs.

Problems

No face recognition capability
Ease of use
Multi-platform
Speed

Solutions

AR and AI for Emotion detection.
Comes with ready prefabs and demo scenes.
Write once and publish directly to desktop and mobile environments.
The tracker parts are low-level C++ libraries, made specifically for 32 and 64-bit devices.

BENEFITS

Precision

  • high end Virtual Try On for Cosmetics
  • large rotations (> 80°)
  • respect of face morphology
  • detection of subtle face characteristics for accurate personalization
  • robust to occlusions
  • robust to back light sources & white backgrounds

Resources

  • Mobile first, ultra optimized code for embedded devices
  • Adaptive to power of smartphone (memory, CPU & GPU) for best trade offs
  • Long battery life
  • Works in very dark environments (>> Snapchat)

Openness

  • Broadest range of platforms, mobile, web, desktop, server, embedded IoT
  • Easy to integrate with source code of reference Apps
  • Privacy protection built in