3D Shape Synthesis

Here we will introduce the 3D new shape synthesis technology developed by our company, a pioneer in the world.

What is 3D Shape Synthesis?

3D new shape synthesis means that a properly trained 3D AI model blends feature vectors extracted from multiple shapes and automatically generates new shapes. We have developed a 3D AI model that can synthesize new 3D shapes.

Mechanism of 3D Car Shape Synthesis

For example, you can create a new car body by blending their tastes well with multiple existing car bodies.

For the new shape synthesis, a network using Variational Autoencoder (VAE), a type of deep learning, is built to create new car body data, with multiple timings for adding feature expressions to the Decoder part of the VAE system to flavor the generated results.

First, get the features from each of the two body meshes. When creating a new car body, you can change the composition ratio and create a car body with a new taste.

Mechanism of 3D Furniture Shape Synthesis Model

Our furniture synthesis model utilizes cutting-edge AI techniques to capture and reconstruct furniture designs. We employ a sophisticated architecture that integrates multiple AI components, including an additional AI specifically for mesh reconstruction. This ensures the creation of resonable and accurate 3D mesh models.

We use a Variational Autoencoder (VAE) architecture, similar to those used in car synthesis, but with distinct encoder and decoder designs. By applying a reparameterization technique, we control the distribution of the latent space, which compresses information about the 3D shape features into a designed distribution.

Here’s a detailed explanation:

1. Encoder : We utilize an advanced encoder to extract detailed features of the furniture, capturing its intricate design elements and structural characteristics.
2. Reparametize : This technique controls the distribution of the latent space after the encoding process. It is a key component of the VAE architecture, ensuring that the latent space follows a designed distribution.
3. Decoder : decoder processes the information from the latent space and converts it into features that are used for mesh reconstruction.
4. Mesh Reconstruction : we implement mesh reconstruction techniques. This allows the model to rebuild the furniture design from the decoded output freatures, ensuring a high level of detail and giving some smoothing and removing outlier in the mesh reconstruction output.

By combining advanced feature extraction techniques with sophisticated decoding for 3D model and mesh reconstruction AI models, our furniture synthesis system achieves resonable performance in generating new furniture designs.

ARIES 3D Synthesis – Astraea Software Product

To gain a deeper understanding of our cutting-edge technology, we invite you to explore our product development in 3D synthesis models. Our innovative solutions are designed to revolutionize various industries by providing advanced capabilities in 3D modeling and simulation. Discover how our technology can enhance your projects and streamline your processes. For more detailed information and to see our 3D synthesis models in action, click the button below: