Speeding-up Object Detection Training for Robotics with FALKON

TL;DR

This paper introduces a new object detection pipeline for robotics that significantly speeds up training by combining deep features with the FALKON kernel method, achieving a 60x faster training time while maintaining performance.

Contribution

The novel pipeline integrates deep region proposals with FALKON for rapid training on large-scale, imbalanced datasets in robotic object detection tasks.

Findings

Achieves 60x faster training speed compared to traditional methods.

Maintains comparable detection performance on standard datasets.

Demonstrates effectiveness in real robotic scenarios with iCubWorld dataset.

Abstract

Latest deep learning methods for object detection provide remarkable performance, but have limits when used in robotic applications. One of the most relevant issues is the long training time, which is due to the large size and imbalance of the associated training sets, characterized by few positive and a large number of negative examples (i.e. background). Proposed approaches are based on end-to-end learning by back-propagation [22] or kernel methods trained with Hard Negatives Mining on top of deep features [8]. These solutions are effective, but prohibitively slow for on-line applications. In this paper we propose a novel pipeline for object detection that overcomes this problem and provides comparable performance, with a 60x training speedup. Our pipeline combines (i) the Region Proposal Network and the deep feature extractor from [22] to efficiently select candidate RoIs and encode them into powerful representations, with (ii) the FALKON [23] algorithm, a novel kernel-based method that allows fast training on large scale problems (millions of points). We address the size and imbalance of training data by exploiting the stochastic subsampling intrinsic into the method and a novel, fast, bootstrapping approach. We assess the effectiveness of the approach on a standard Computer Vision dataset (PASCAL VOC 2007 [5]) and demonstrate its applicability to a real robotic scenario with the iCubWorld Transformations [18] dataset.