Real-time Video Analysis AI Engine
The real-time AI video analysis is used to understand the emotional atmosphere of the scene based on the flow of the content. The emotional atmosphere helps to recognize the emotion of the viewer, and mixed with the AI analysis of the scene’s various elements such as character, environment, color, and sound, the AI finds the emotion value that best matches the scene and story.
Edge Device Equipped with Emotion Analysis AI Model
The edge device (JETSON NANO) is equipped with the emotion information analysis technology, allowing for a lightweight model and fast inference.
Emotion Analysis AI Model
Using vectors such as image (CLIP), audio (Hubert), color, and intensity, the AI model analyzes 7 types of emotions in real-time.
Emotion AI Sound System
It provides sound tuning matching the emotion classes that have been selected through video emotion AI analysis. It elevates the emotional atmosphere of the video and creates an immersive environment by enhancing the sounds of characters, objects, and environment. It is a unique real-time emotion sound effect that is different from genre-based sound tuning.
Sound tuning that refines the atmosphere of video, balances the sound of object and environment, and elevates spatial depth. Reaches the desired sound effect by reflecting preset sound tuning values to elements such as Dialog / 3D Immersive / Pitch / Volume / EQ