iOS App File Layout

-- First layer
SemanticSegmentation-CoreML/
    -- Second layer
    Main.storyboard
    AppDelegate.swift
    LiveImageViewController.swift
    LiveMetalCameraViewController.swift
    StillImageViewController.swift
    DrawingSegmentationView.swift
    LiveFaceDetectionAndFaceParsingViewController.swift
    SegmentationResultMLMultiArray.swift
    Measure.swift
    VideoCapture.swift
    MetalCamera/
        -- Third layer
        MetalRenderingDevice.swift
        MetalVideoView.swift
        CameraTextureGenerater.swift
        SegmentationTextureGenerater.swift
        MultitargetSegmentationTextureGenerater.swift
        OverlayingTexturesGenerater.swift
        CroppedTextureGenerater.swift
        MaskTextureGenerater.swift
        Texture.swift
        Shaders/
            -- Fourth layer
            Shaders.metal
        Utils/
            -- Fourth layer
            Maths.swift
        mlmodel/
            -- Fourth layer
            DeepLabV3.mlmodel
            DeepLabV3FP16.mlmodel
            DeepLabV3Int8LUT.mlmodel
            FaceParsing.mlmodel

CoreMl Models

Segmentation Data:

Segmentation shaders take buffer data that indicates classification results (like SegmentationValue) and use these to apply colors conditionally.

Layered Rendering:

Different shaders (like segmentation_render_target and multitarget_segmentation_render_target) suggest that the application might handle layered rendering where each layer represents different data or visual aspects. These could be composited together to form the final image, each layer possibly using different shaders for specialized rendering based on the layer's purpose.

Reference

GitHub - likedan/Awesome-CoreML-Models: Largest list of models for Core ML (for iOS 11+)

https://github.com/juanmorillios/List-CoreML-Models

GitHub - satoshi0212/MSLExamples: Metal Shader Language(MSL) examples.

https://github.com/tucan9389/SemanticSegmentation-CoreML

Shadertoy