Inferring 3D Object Pose in RGB-D Images

TL;DR

This paper presents a CNN-based method for inferring 3D object poses in RGB-D images, enabling accurate object replacement with a significant performance boost over existing techniques.

Contribution

The work introduces a novel CNN trained on synthetic data for pose prediction, improving accuracy and speed in 3D object detection in RGB-D scenes.

Findings

48% improvement in 3D detection performance

Outperforms state-of-the-art methods

Faster processing time

Abstract

The goal of this work is to replace objects in an RGB-D scene with corresponding 3D models from a library. We approach this problem by first detecting and segmenting object instances in the scene using the approach from Gupta et al. [13]. We use a convolutional neural network (CNN) to predict the pose of the object. This CNN is trained using pixel normals in images containing rendered synthetic objects. When tested on real data, it outperforms alternative algorithms trained on real data. We then use this coarse pose estimate along with the inferred pixel support to align a small number of prototypical models to the data, and place the model that fits the best into the scene. We observe a 48% relative improvement in performance at the task of 3D detection over the current state-of-the-art [33], while being an order of magnitude faster at the same time.