Abstract: We present a technique for three-dimensional (3D) object reconstruction utilizing an ultrasonic array sensor and a variational autoencoder (VAE) within a high-interference environment.

Variational Autoencoders (VAEs) address some of the limitations of traditional autoencoders by introducing a probabilistic approach to encoding and decoding. The motivation behind VAEs lies in their ability to generate new data samples by sampling from a learned distribution in the latent space rather than from a latent vector as was the case ...

Jan 25, 2021 · In a harsh remote collaboration setting, data compression techniques such as autoencoder can be utilized to obtain and transmit the data in terms of latent variables in a compact form.

Sep 17, 2021 · In this paper, we propose voxel-based 3D object reconstruction (V3DOR) from a single 2D image for better accuracy, one using autoencoders (AE) and another using variational autoencoders (VAE).

Mar 19, 2024 · In this study, we focus on the 3D-VAE-GAN models, a novel combination of generative adversarial networks (GANs) and variational autoencoders (VAEs) in the field of 3D object generation from 2D images.

Jan 25, 2021 · Based on the experiments on the ModelNet and Pascal3D datasets, the proposed approach shows consistently superior performance over autoencoder and variational autoencoder up to 70% data loss.

In this paper, we propose a novel framework called mesh variational autoencoders (mesh VAE), which leverages the power of neural networks to explore the latent space behind deforming 3D shapes, and is able to generate new models not existing in the original dataset.

This document explains how to use the 3D Variational Autoencoder (VAE) for encoding 3D assets into the point cloud-structured latent space and reconstructing them. The VAE is a fundamental component of the GaussianAnything framework that bridges 3D asset representation and the generation pipeline. The primary purposes of VAE reconstruction are:

Based on the experiments on the ModelNet and Pascal3D datasets, the proposed approach shows consistently superior performance over autoencoder and variational autoencoder up to 70% data loss.

Oct 27, 2023 · To tackle this problem, we propose a Scene Graph Masked Variational Auto-Encoder (SG-MVAE) framework that fully captures the relationships between objects to generate more realistic 3D scenes. Specifically, we first introduce a relationship completion module that adaptively learns the missing relationships between objects in the scene graph.