HomeIoT ModelsMediaPipe-Selfie-Segmentation

    MediaPipe-Selfie-Segmentation

    Segments the person from background in a selfie image and realtime background segmentation in video conferencing.

    Light-weight model that segments a person from the background in square or landscape selfie and video conference imagery.

    Qualcomm® QCS8550
    QCS8550 (Proxy)
    0.79ms
    Inference Time
    0-2MB
    Memory Usage
    118NPU
    Layers

    Technical Details

    Model checkpoint:Square
    Input resolution (Square):256x256
    Input resolution (Landscape):144x256
    Number of parameters:106K
    Model size:454 KB

    Applicable Scenarios

    • Camera
    • Instant Photo Studio
    • Video Conferencing
    • Personalized Marketing Content
    • Interactive Gaming Avatar
    • Real-time Portrait Editor
    • Dynamic Wallpaper Generator

    Licenses

    Source Model:APACHE-2.0
    Deployable Model:AI Model Hub License

    Supported IoT Devices

    • QCS8550 (Proxy)

    Supported IoT Chipsets

    • Qualcomm® QCS8550