An empirical study of context in object detection

By Santosh Kumar Divvala, Derek Hoiem, James Hays, Alexei A. Efros, Martial Hebert. 2009

606 citations recorded by the source corpus.

This paper presents an empirical evaluation of the role of context in a contemporary, challenging object detection task - the PASCAL VOC 2008. Previous experiments with context have mostly been done on home-grown datasets, often with non-standard baselines, making it difficult to isolate the contribution of contextual information. In this work, we present our analysis on a standard dataset, using top-performing local appearance detectors as baseline. We evaluate several different sources of context and ways to utilize it. While we employ many contextual cues that have been used before, we also propose a few novel ones including the use of geographic context and a new approach for using object spatial support.

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): “Object relation in post-processing Most early works use object relations as a post-processing step [12, 17, 46, 47, 36, 17].”
Inside-Outside Net: Detecting Objects in Context with Skip Pooling and Recurrent Neural Networks (2015, neutral): “Global average pooling provides information about the entire image, similar to the features used for scene or image classification [25, 20].”
Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation (2015, methodology): “outperforms the published state-of-the-art using the same training dataset. II. RELATED WORK Exploiting contextual information has been widely studied in the literature, e.g., the work in [30], [17], [5], [6]. For example, the early work “TAS” [17] models different types of spatial context between T”
Gather-Excite: Exploiting Feature Context in Convolutional Neural Networks (2018, background): “While the term “context” is overloaded [6], in this work we focus specifically on feature context, namely the information captured by the feature extractor responses (i.”
Don't just listen, use your imagination: Leveraging visual common sense for non-visual tasks (2015, background): “One instantiation of visual common sense that has been leveraged in the vision community in the past is the use of contextual reasoning for improved recognition [22, 12, 21, 17, 25, 50].”
Neural Motifs: Scene Graph Parsing with Global Context (2017, background): “Context Many methods have been proposed for modeling semantic context in object recognition [7].”
DeepID-Net: multi-stage and deformable deep convolutional neural networks for object detection (2014, background): “The context information inves tigated in previous works includes regions surrounding objects [9, 12, 22], object-scene interaction [ 13], and the presence, location, orientation and size relationship among ob jects [3, 57, 58, 11, 41, 22, 49, 13, 61, 12, 59, 40, 10, 45, 51].”
OCNet: Object Context Network for Scene Parsing (2018, methodology): “[5] used to conduct an empirical study of the influence of the context in object detection.”
Modeling Visual Context is Key to Augmenting Object Detection Datasets (2018, background): “For this reason, a taxonomy of contextual information is proposed in [2] to better understand what type of visual context is useful for object detection.”
DeepID-Net: Object Detection with Deformable Part Based Convolutional Neural Networks (2017, neutral): “object-scene interaction [60], [61], and the presence, location, orientation and size relationship among objects [33], [58], [59], [60], [62], [63], [64], [65], [66], [67], [68], [69], [70], [71], [72].”
Context tracker: Exploring supporters and distracters in unconstrained environments (2011, positive): “Context information has been applied actively in object detection [9], object classification [20, 23], object recognition [25].”
Predicting occupation via human clothing and contexts (2011, neutral): “[8].”
Situation Recognition: Visual Semantic Role Labeling for Image Understanding (2016, neutral): “Many methods have been proposed for modeling semantic context in activity recognition [6].”
Recent advances and trends in visual tracking: A review (2011, background): “A recent detailed study of context can be found in [113], which compared several kinds of context information within the object detection area.”
Symbiotic Segmentation and Part Localization for Fine-Grained Categorization (2013, background): “It is well known [13] that at the base category level the background often provides valuable context for categorization.”
Superparsing (2010, background): “The literature contains diverse proposed image parsing methods, including ones that estimate labels pixel by pixel [1,2], ones that aggregate features over segmentation regions [3,4,5,6], and ones that predict object bounding boxes [7,8,9,10].”
Nonparametric Scene Parsing via Label Transfer (2011, methodology): “For information on obtaining reprints of this article, please send e-mail to: tpami@computer.org, and reference IEEECS Log Number TPAMISI-2010-08-0663.”
Region-Based Saliency Detection and Its Application in Object Recognition (2014, background): “In practice, in some cases, background may be useful to the object recognition [16], [53].”