gradio-sync3dcompare 0.0.8

Synchronized side-by-side 3D comparison viewer for Gradio — supports PLY and GLB files with shared camera control.

gradio_sync3dcompare

Synchronized side-by-side 3D comparison viewer for Gradio — supports PLY and GLB files with shared camera control.

Installation

pip install gradio_sync3dcompare

Usage

import gradio as gr
from gradio_sync3dcompare import Sync3DCompare
import os

SAMPLE_DIR = os.path.join(os.path.dirname(__file__), "sample_data")

def get_path(filename):
    return os.path.join(SAMPLE_DIR, filename)


with gr.Blocks(title="Sync3DCompare Demo") as demo:
    gr.Markdown("# Sync3DCompare — Synchronized 3D Comparison Viewer")
    gr.Markdown(
        "Compare 3D reconstruction outputs side-by-side with synchronized orbit, pan, and zoom. "
        "Rotate one view and all others follow."
    )

    with gr.Tabs():

        # --- Tab A: PLY comparison ---
        with gr.TabItem("PLY Comparison"):
            gr.Markdown("### Three PLY point clouds compared side-by-side")
            Sync3DCompare(
                label="PLY Comparison",
                value=[
                    {
                        "name": "Ground Truth",
                        "path": get_path("ground_truth.ply"),
                        "type": "ply",
                        "color": [80, 200, 120],
                    },
                    {
                        "name": "Method A",
                        "path": get_path("method_a.ply"),
                        "type": "ply",
                        "color": [100, 160, 240],
                    },
                    {
                        "name": "Method B",
                        "path": get_path("method_b.ply"),
                        "type": "ply",
                        "color": [240, 140, 80],
                    },
                ],
                render_mode="points",
                sync_camera=True,
                point_size=2.0,
                height=500,
            )

        # --- Tab B: GLB comparison ---
        with gr.TabItem("GLB Comparison"):
            gr.Markdown("### Two GLB mesh outputs compared in native mode")
            Sync3DCompare(
                label="GLB Comparison",
                value=[
                    {
                        "name": "Mesh A (detailed)",
                        "path": get_path("mesh_a.glb"),
                        "type": "glb",
                    },
                    {
                        "name": "Mesh B (coarse)",
                        "path": get_path("mesh_b.glb"),
                        "type": "glb",
                    },
                ],
                render_mode="native",
                sync_camera=True,
                point_size=2.0,
                height=500,
            )

        # --- Tab C: Mixed PLY + GLB ---
        with gr.TabItem("Mixed Format (Points Mode)"):
            gr.Markdown(
                "### PLY + GLB compared in `points` mode — GLB is sampled to point cloud for fair comparison"
            )
            Sync3DCompare(
                label="Mixed Format Comparison",
                value=[
                    {
                        "name": "PLY Point Cloud",
                        "path": get_path("ground_truth.ply"),
                        "type": "ply",
                        "color": [80, 200, 120],
                    },
                    {
                        "name": "GLB as Points",
                        "path": get_path("mesh_a.glb"),
                        "type": "glb",
                    },
                ],
                render_mode="points",
                sync_camera=True,
                point_size=3.0,
                height=500,
            )


if __name__ == "__main__":
    demo.launch()

Sync3DCompare

Initialization

name	type	default	description
value	list[dict]| None	value = None	None
label	str| None	value = "3D Comparison"	None
render_mode	"points"| "native"	value = "points"	None
sync_camera	bool	value = True	None
point_size	float	value = 2.0	None
height	int	value = 500	None
num_views	int	value = 2	None
interactive	bool	value = True	None
visible	bool	value = True	None
elem_id	str| None	value = None	None
elem_classes	list[str]| str| None	value = None	None
render	bool	value = True	None

Events

name	description
change

User function

The impact on the users predict function varies depending on whether the component is used as an input or output for an event (or both).

• When used as an Input, the component only impacts the input signature of the user function.
• When used as an output, the component only impacts the return signature of the user function.

The code snippet below is accurate in cases where the component is used as both an input and an output.

• As input: Should return, the output data received by the component from the user's function in the backend.

def predict(
    value: Unknown
) -> list[dict]| None:
    return value