trellis

R2026b

Generate and modify 3D assets using Trellis model

Since R2026b

expand all in page

Description

Add-On Required: This feature requires the 3D Asset Processing Library for MATLAB add-on.

The trellis object uses the Trellis model to generate 3D assets from single-view images, multi-view images, or natural language text descriptions, and modify existing 3D assets by using natural language text instructions [1].

Note

This feature requires Deep Learning Toolbox™ and a GPU with at least 16 GB of memory and a CUDA® compute capability 7.0 or higher.

Creation

Syntax

trellisObj = trellis
trellisObj = trellis(ShowProgressBar=showProgressBar)

Description

trellisObj = trellis creates a trellis object, trellisObj, for generating and modifying 3D assets.

example

trellisObj = trellis(ShowProgressBar=showProgressBar) specifies whether to display the progress bar during object creation.

Input Arguments

expand all

Display progress bar during object creation, specified as a logical 1 (true) or 0 (false). By default, the trellis function displays a progress bar while it initializes the object and sets up the model.

Output Arguments

expand all

Trellis 3D asset generator, returned as a trellis object.

Object Functions

generateFromImageGenerate 3D asset from one or more images
generateFromTextGenerate 3D asset from text prompt
modifyModify existing 3D asset using text prompt
releaseGPUMemoryMove models to CPU and free GPU memory

Examples

collapse all

This example uses:

Open Live Script

Note: This example requires Deep Learning Toolbox and a GPU with at least 16 GB of memory and a CUDA compute capability of 7.0 or higher.

Create a trellis object.

trellisObj = trellis;

Read a single RGB image into the workspace.

img = imread("ObjaverseAssetImage.jpeg");
imshow(img)

Generate a 3D asset from the image by using the generateFromImage object function of the trellis object.

meshObj = generateFromImage(trellisObj,img);

Display the generated 3D asset.

show(meshObj)

Generate a 3D asset from the image with custom generation parameters. Return the preprocessing information info, enabling you to inspect the image after background removal.

[meshObj_custom,info] = generateFromImage(trellisObj,img,GeometryGenerationGuidanceStrength=8.5,GeometryGenerationSamplingSteps=50,...
    TextureGenerationGuidanceStrength=3.5,TextureGenerationSamplingSteps=50,SimplifyRatio=0.9,TextureSize=2048,Seed=42);

Display the 3D asset generated using the custom parameters.

show(meshObj_custom)

Display the preprocessed image to inspect the background removal applied before asset generation.

figure
imshow(info.PreprocessedImage)
title("Preprocessed Image")

This example uses:

Open Live Script

Note: This example requires Deep Learning Toolbox and a GPU with at least 16 GB of memory and a CUDA compute capability of 7.0 or higher.

Create a trellis object.

trellisObj = trellis;

Generate a 3D asset from a text description by using the generateFromText object function of the trellis object.

meshObj = generateFromText(trellisObj,"A red sedan car");

Display the generated 3D asset.

show(meshObj)

Generate a 3D asset using a detailed text description and custom generation parameters to produce a higher-quality mesh.

meshObj_custom = generateFromText(trellisObj,"A red sedan car with black roof",GeometryGenerationGuidanceStrength=9,...
    GeometryGenerationSamplingSteps=50,TextureGenerationGuidanceStrength=8,TextureGenerationSamplingSteps=50,...
    SimplifyRatio=0.9,TextureSize=1024,Seed=42);

Display the 3D asset generated using the custom parameters.

show(meshObj_custom)

Limitations

macOS does not support generating 3D mesh assets from images or text prompts.

References

[1] Xiang, Jianfeng, Zelong Lv, Sicheng Xu, et al. “Structured 3D Latents for Scalable and Versatile 3D Generation.” Preprint, arXiv, 2024. https://doi.org/10.48550/ARXIV.2412.01506.

Version History

Introduced in R2026b

See Also

| | | | | | | |

Topics