摘要三维重建在视觉方面具有很高的研究价值, 在机器人视觉导航, 智能车环境感知系统以及
虚拟现实中被广泛应用. 本文对近年来国内外基于视觉的三维重建方法的研究工作进行了总结和 …

We study the notion of consistency between a 3D shape and a 2D observation and propose
a differentiable formulation which allows computing gradients of the 3D shape given an …

We present 3DMV, a novel method for 3D semantic scene segmentation of RGB-D scans
using a joint 3D-multi-view prediction network. In contrast to existing methods that either use …

We present a framework for learning single-view shape and pose prediction without using
direct supervision for either. Our approach allows leveraging multi-view observations from …

Planar primitives tend to be incorrectly detected or incomplete in complex scenes where
adhesions exist between different objects, resulting in topology errors in the reconstructed …

In this paper, we present a learning based approach to depth fusion, ie, dense 3D
reconstruction from multiple depth images. The most common approach to depth fusion is …

The efficient fusion of depth maps is a key part of most state-of-the-art 3D reconstruction
methods. Besides requiring high accuracy, these depth fusion methods need to be scalable …

Supervised 3D reconstruction has witnessed a significant progress through the use of deep
neural networks. However, this increase in performance requires large scale annotations of …

We present a novel online depth map fusion approach that learns depth map aggregation in
a latent feature space. While previous fusion methods use an explicit scene representation …

In this paper, we consider the problem of reconstructing a dense 3D model using images
captured from different views. Recent methods based on convolutional neural networks …