# Large Scale Novel Object Discovery in 3D

## TL;DR

This paper introduces a novel 3D object discovery method that leverages supervoxels and deep learning to identify unseen objects in point cloud data, demonstrating effective clustering and generalization from limited training data.

## Contribution

The paper proposes a new approach combining supervoxels and a Siamese network for discovering novel objects in 3D point clouds, trained on few models and tested on unseen objects.

## Key findings

- Successfully discovers numerous unseen objects in 3D point clouds.
- Achieves high accuracy in clustering unknown objects based on learned embeddings.
- Demonstrates strong generalization from limited training data.

## Abstract

We present a method for discovering never-seen-before objects in 3D point clouds obtained from sensors like Microsoft Kinect. We generate supervoxels directly from the point cloud data and use them with a Siamese network, built on a recently proposed 3D convolutional neural network architecture. We use known objects to train a non-linear embedding of supervoxels, by optimizing the criteria that supervoxels which fall on the same object should be closer than those which fall on different objects, in the embedding space. We test on unknown objects, which were not seen during training, and perform clustering in the learned embedding space of supervoxels to effectively perform novel object discovery. We validate the method with extensive experiments, quantitatively showing that it can discover numerous unseen objects while being trained on only a few dense 3D models. We also show very good qualitative results of object discovery in point cloud data when the test objects, either specific instances or even categories, were never seen during training.