What question did this study set out to answer?

This research aims to enhance object detection accuracy using region-based convolutional networks.

May 25, 2015

Region-Based Convolutional Networks for Accurate Object Detection and Segmentation

Key Points

This research aims to enhance object detection accuracy using region-based convolutional networks.
Utilized high-capacity convolutional networks for processing bottom-up region proposals.
Implemented supervised pre-training on an auxiliary task followed by domain-specific fine-tuning.
Achieved mean average precision assessment using canonical PASCAL VOC Challenge datasets.
Achieved a mean average precision of 62.4 percent on VOC 2012.
Improved mean average precision by more than 50 percent relative to previous best results.

Abstract

Object detection performance, as measured on the canonical PASCAL VOC Challenge datasets, plateaued in the final years of the competition. The best-performing methods were complex ensemble systems that typically combined multiple low-level image features with high-level context. In this paper, we propose a simple and scalable detection algorithm that improves mean average precision (mAP) by more than 50 percent relative to the previous best result on VOC 2012-achieving a mAP of 62.4 percent. Our approach combines two ideas: (1) one can apply high-capacity convolutional networks (CNNs) to bottom-up region proposals in order to localize and segment objects and (2) when labeled training data are scarce, supervised pre-training for an auxiliary task, followed by domain-specific fine-tuning, boosts performance significantly. Since we combine region proposals with CNNs, we call the resulting model an R-CNN or Region-based Convolutional Network. Source code for the complete system is available at http://www.cs.berkeley.edu/~rbg/rcnn.

Demander à l'IA

Bookmark