Using Local Context To Improve Face Detection

Kruppa, Hannes; Schiele, Bernt

doi:10.5244/c.17.5

Cited by 28 publications

(18 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Spatial context is incorporated from inter-pixel statistics [7,13,15,20,24,27,30,34,37,38] and from pairwise relations between regions in images [6,11,16,19,31].…”

Section: Spatial Contextmentioning

confidence: 99%

“…Local context features can capture different local relations such as pixel, region and object interactions. Many object categorization models have used local context from pixels [6,7,13,15,30], patches [16,19,31] and objects [11,12,25,33,35,38] that surrounds the target object, greatly improving the task of object categorization. These interactions are reviewed in detail in Section 4.…”

Section: Local Contextmentioning

confidence: 99%

“…Kruppa and Schiele [15] investigated the role of local context for face detection algorithms. In their work, an appearance-based object detector is trained with instances that contain a persons entire head, neck and part of the upper body (as shown in Figure 6 (a)).…”

Section: Local Contextmentioning

confidence: 99%

See 2 more Smart Citations

Context based object categorization: A critical survey

Galleguillos

Belongie

2010

Computer Vision and Image Understanding

305

164

View full text Add to dashboard Cite

Abstract. The goal of object categorization is to locate and identify instances of an object category within an image. Recognizing an object in an image is difficult when images present occlusion, poor quality, noise or background clutter, and this task becomes even more challenging when many objects are present in the same scene. Several models for object categorization use appearance and context information from objects to improve recognition accuracy. Appearance information, based on visual cues, can successfully identify object classes up to a certain extent. Context information, based on the interaction among objects in the scene or on global scene statistics, can help successfully disambiguate appearance inputs in recognition tasks. In this work we review different approaches of using contextual information in the field of object categorization and discuss scalability, optimizations and possible future approaches.

show abstract

“…Spatial context is incorporated from inter-pixel statistics [7,13,15,20,24,27,30,34,37,38] and from pairwise relations between regions in images [6,11,16,19,31].…”

Section: Spatial Contextmentioning

confidence: 99%

Section: Local Contextmentioning

confidence: 99%

See 1 more Smart Citation

Context based object categorization: A critical survey

Galleguillos

Belongie

2010

Computer Vision and Image Understanding

305

164

View full text Add to dashboard Cite

show abstract

“…Face detection using body information has been discussed in some early works [16] [19], but how to use these information in unconstraint setting is still unclear. Recently Pascal VOC person layout competition aims to predict the location of head, hands and feet given the location of the body [5].…”

Section: A Related Workmentioning

confidence: 99%

Structural models for face detection

Yan

Zhang

Lei

et al. 2013

2013 10th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG)

View full text Add to dashboard Cite

Abstract-Despite the success in the last two decades, the state-of-the-art face detectors still have problems in dealing with images in the wild for the large appearance variations. Instead of taking appearance variations as black boxes and leaving them to statistical learning algorithms, we propose a structural face model to explicitly represent them. Our hierarchical part based structural face model enables part subtype option to describe appearance variations of the local part, and part deformation to capture the deformable variations between different poses and expressions. In the process of detection, the input candidate is first fitted by the structural model to infer the part location and part subtype, and the confidence score is then computed based on the fitted configuration to reduce the influence of structure variation. Besides the face model, we utilize the co-occurrence of face and body to further boost the face detection performance. We present a method for training phrase based body detectors, and propose a structural context model to jointly use the results of face detector and various body detectors. Experiments on the challenging FDDB show that our method has state-of-the-art performance, compared with other commercial and academic systems.

show abstract

“…Our own work targets object localization in images, a topic for which context has also successfully been applied: Kruppa and Schiele use a fixed region surrounding a detection window to improve the detection of face with very low resolution [17]. Dalal and Triggs showed that one achieves better results in pedestrian detection if a detection window larger than the actual person is used [9].…”

Section: Introductionmentioning

confidence: 99%

Object Localization with Global and Local Context Kernels

Blaschko

Lampert

2009

Procedings of the British Machine Vision Conference 2009

View full text Add to dashboard Cite

Recent research has shown that the use of contextual cues significantly improves performance in sliding window type localization systems. In this work, we propose a method that incorporates both global and local context information through appropriately defined kernel functions. In particular, we make use of a weighted combination of kernels defined over local spatial regions, as well as a global context kernel. The relative importance of the context contributions is learned automatically, and the resulting discriminant function is of a form such that localization at test time can be solved efficiently using a branch and bound optimization scheme. By specifying context directly with a kernel learning approach, we achieve high localization accuracy with a simple and efficient representation. This is in contrast to other systems that incorporate context for which expensive inference needs to be done at test time. We show experimentally on the PASCAL VOC datasets that the inclusion of context can significantly improve localization performance, provided the relative contributions of context cues are learned appropriately.

show abstract

Using Local Context To Improve Face Detection

Cited by 28 publications

References 13 publications

Context based object categorization: A critical survey

Context based object categorization: A critical survey

Structural models for face detection

Object Localization with Global and Local Context Kernels

Contact Info

Product

Resources

About