AttentionNet: Aggregating Weak Directions for Accurate Object Detection

Abstract

We present a novel detection method using a deep convolutional neural network (CNN), named AttentionNet. We cast an object detection problem as an iterative classification problem, which is the most suitable form of a CNN. AttentionNet provides quantized weak directions pointing a target object and the ensemble of iterative predictions from AttentionNet converges to an accurate object boundary box. Since AttentionNet is a unified network for object detection, it detects objects without any separated models from the object proposal to the post bounding-box regression. We evaluate AttentionNet by a human detection task and achieve the state-of-the-art performance of 65% (AP) on PASCAL VOC 2007/2012 with an 8-layered architecture only.

Keywords

Pascal (unit)Object detectionMinimum bounding boxComputer scienceConvolutional neural networkArtificial intelligenceObject (grammar)Bounding overwatchPattern recognition (psychology)Iterative methodClassifier (UML)Cognitive neuroscience of visual object recognitionDeep learningArtificial neural networkComputer visionAlgorithmImage (mathematics)

Affiliated Institutions

Related Publications

Is object localization for free? - Weakly-supervised learning with convolutional neural networks

Maxime Oquab , Léon Bottou , Ivan Laptev +1 more

Successful methods for visual object recognition typically rely on training datasets containing lots of richly annotated images. Detailed image annotation, e.g. by object boundi...

2015 915 citations

Object Detection via a Multi-region and Semantic Segmentation-Aware CNN Model

Spyros Gidaris , Nikos Komodakis

We propose an object detection system that relies on a multi-region deep convolutional neural network (CNN) that also encodes semantic segmentation-aware features. The resulting...

2015 799 citations

Deep Neural Networks for Object Detection

Christian Szegedy , Alexander Toshev , Dumitru Erhan

Deep Neural Networks (DNNs) have recently shown outstanding performance on image classification tasks [14]. In this paper we go one step further and address the problem of objec...

2013 1200 citations

Fast R-CNN

Ross Girshick

This paper proposes a Fast Region-based Convolutional Network method (Fast R-CNN) for object detection. Fast R-CNN builds on previous work to efficiently classify object proposa...

2015 2015 IEEE International Conference on... 26511 citations

Fast R-CNN

Ross Girshick

This paper proposes a Fast Region-based Convolutional Network method (Fast R-CNN) for object detection. Fast R-CNN builds on previous work to efficiently classify object proposa...

2015 arXiv (Cornell University) 1766 citations

Publication Info

Year: 2015
Type: article
Pages: 2659-2667
Citations: 178
Access: Closed

External Links

View on DOI.org

Social Impact

Altmetric

AttentionNet: Aggregating Weak Directions for Accurate Object Detection

PlumX Metrics

Social media, news, blog, policy document mentions

Citation Metrics

178

OpenAlex

Cite This

APA Style

                            
                                    Donggeun Yoo, 
                                
                                    Sunggyun Park, 
                                
                                    Joon‐Young Lee
                                
                                et al.
                            
                            (2015). 
                            AttentionNet: Aggregating Weak Directions for Accurate Object Detection. 
                            
                            , 2659-2667.
                            https://doi.org/10.1109/iccv.2015.305

Identifiers

DOI: 10.1109/iccv.2015.305