Attentive Contexts for Object Detection

By Jianan Li, Yunchao Wei, Xiaodan Liang, Jian Dong, Tingfa Xu, Jiashi Feng, Shuicheng Yan. 2016

234 citations recorded by the source corpus.

Modern deep neural network-based object detection methods typically classify candidate proposals using their interior features. However, global and local surrounding contexts that are believed to be valuable for object detection are not fully exploited by existing methods yet. In this work, we take a step towards understanding what is a robust practice to extract and utilize contextual information to facilitate object detection in practice. Specifically, we consider the following two questions: “how to identify useful global contextual information for detecting a certain object?” and “how to exploit local context surrounding a proposal for better inferring its contents?” We provide preliminary answers to these questions through developing a novel attention to context convolution neural network (AC-CNN)-based object detection model. AC-CNN effectively incorporates global and local contextual information into the region-based CNN (e.g., fast R-CNN and faster R-CNN) detection framework and provides better object detection performance. It consists of one attention-based global contextualized (AGC) subnetwork and one multi-scale local contextualized (MLC) subnetwork. To capture global context, the AGC subnetwork recurrently generates an attention map for an input image to highlight useful global contextual locations, through multiple stacked long short-term memory layers. For capturing surrounding local context, the MLC subnetwork exploits both the inside and outside contextual information of each specific proposal at multiple scales. The global and local context are then fused together for making the final decision for detection. Extensive experiments on PASCAL VOC 2007 and VOC 2012 well demonstrate the superiority of the proposed AC-CNN over well-established baselines.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Relation Networks for Object Detection (2017, background): “It takes variable number of inputs, runs in parallel (as opposed to sequential relation modeling [29, 44, 6]), is fully differen-”
MegDet: A Large Mini-Batch Object Detector (2017, methodology): “Based on our MegDet, we integrate the techniques including OHEM [35], atrous convolution [40, 2], stronger base models [38, 18], large kernel [28], segmentation supervision [27, 34], diverse network structure [12, 32, 36], contextual modules [22, 9], ROIAlign [14] and multi-scale training and testin”
Spatial Memory for Context Reasoning in Object Detection (2017, background): “It has resemblance to spatial attention, which has been explored in many high-level tasks [47, 68, 91, 92].”
Right for the Right Reason: Training Agnostic Networks (2018, background): “telling you that it is unlikely to be an elephant inside a telephone box), as shown in many recent works [3], [13], [14].”
Scale-Aware Fast R-CNN for Pedestrian Detection (2015, background): “Deep Learning Model: Convolutional neural networks (CNNs) have recently been successfully applied in generic object recognition [15], [16], [21], [23], [26], [28], [28], [40], [40], [41], [43], [46], [52], [63].”
Recent Advances in Deep Learning for Object Detection (2019, background): “code contextual information is to implicitly encode region features by adding image features surrounding region proposals and a large number of approaches have been proposed based on this idea [131, 106, 140, 141, 142, 143].”
Extended Feature Pyramid Network for Small Object Detection (2020, background): “Besides, directly adding context attention [23] is also an effective way to enhance detectors’ performance.”
Object Detection in 20 Years: A Survey (2019, unknown): “The second method is to think of the global context as a kind of sequential information and to learn it with the recurrent neural networks [86], [98].”
Adaptive Convolution for Object Detection (2019, background): “Since the information of region Ω is pooled into one cell, elements of L can merge more local information containing rich semantic information, which is proved to be helpful for more accurate object detection [30], [31].”
An End-to-End Architecture for Class-Incremental Object Detection with Knowledge Distillation (2019, background): “[9] further explored local and global contextual information for improved object detection.”
PiCANet: Learning Pixel-Wise Contextual Attention for Saliency Detection (2017, background): “[20] utilize attention to attend to the global context to guide object detection.”
SR-LSTM: State Refinement for LSTM Towards Pedestrian Trajectory Prediction (2019, background): “Attention mechanisms have been proven to be signiﬁcantly effective for relevant data selection in various tasks [22,38, 41, 43].”