Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation


Object detection performance, as measured on the canonical PASCAL VOC dataset, has plateaued in the last few years. The best-performing methods are complex ensemble systems that typically combine 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 30% relative to the previous best result on VOC 2012 -- achieving a mAP of 53.3%. Our approach combines two key insights: (1) one can apply high-capacity convolutional neural networks (CNNs) to bottom-up region proposals in order to localize and segment objects and (2) when labeled training data is scarce, supervised pre-training for an auxiliary task, followed by domain-specific fine-tuning, yields a significant performance boost. Since we combine region proposals with CNNs, we call our method R-CNN: Regions with CNN features. We also present experiments that provide insight into what the network learns, revealing a rich hierarchy of image features. Source code for the complete system is available at

DOI: 10.1109/CVPR.2014.81

Extracted Key Phrases

Showing 1-10 of 42 references
Showing 1-10 of 2,081 extracted citations
Citations per Year

3,408 Citations

Semantic Scholar estimates that this publication has received between 3,174 and 3,661 citations based on the available data.

See our FAQ for additional information.