Fine-grained 3D object recognition: an approach and experiments

TL;DR

This paper compares hand-crafted and deep learning approaches for 3D object recognition, demonstrating that deep learning features perform better in open-ended domains and that combining features improves accuracy.

Contribution

It introduces an offline and online 3D object recognition system and evaluates the effectiveness of deep learning versus hand-crafted features in open-ended recognition tasks.

Findings

Deep learning features outperform hand-crafted features in open-ended recognition.

Combining hand-crafted and deep learning features increases classification accuracy.

Deep learning-based approach is more suitable for open-ended 3D object recognition.

Abstract

Three-dimensional (3D) object recognition technology is being used as a core technology in advanced technologies such as autonomous driving of automobiles. There are two sets of approaches for 3D object recognition: (i) hand-crafted approaches like Global Orthographic Object Descriptor (GOOD), and (ii) deep learning-based approaches such as MobileNet and VGG. However, it is needed to know which of these approaches works better in an open-ended domain where the number of known categories increases over time, and the system should learn about new object categories using few training examples. In this paper, we first implemented an offline 3D object recognition system that takes an object view as input and generates category labels as output. In the offline stage, instance-based learning (IBL) is used to form a new category and we use K-fold cross-validation to evaluate the obtained object recognition performance. We then test the proposed approach in an online fashion by integrating the code into a simulated teacher test. As a result, we concluded that the approach using deep learning features is more suitable for open-ended fashion. Moreover, we observed that concatenating the hand-crafted and deep learning features increases the classification accuracy.