Techniques for Still Image Scene Classification and Object Detection

Viitaniemi, Ville; Laaksonen, Jorma

doi:10.1007/11840930_4

Cited by 23 publications

(12 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At this point of operation, approximately 55% of all cats bounding boxes returned are correct and 47% of all dog boxes. At the same time, we correctly localize 50% of method \ data set cat dog ESS w/ bag-of-visual-words kernel 0.223 0.148 Viitaniemi/Laaksonen [23] 0.179 0.131 Shotton/Winn [9] 0.151 0.118 all cats in the dataset and 42% of all dogs. Moving along the curve to the left, only objects are included into the evaluation which have higher confidence scores assigned to them.…”

Section: Pascal Voc 2006 Datasetmentioning

confidence: 78%

Beyond sliding windows: Object localization by efficient subwindow search

Lampert¹,

Blaschko²,

Hofmann

2008

2008 IEEE Conference on Computer Vision and Pattern Recognition

584

458

View full text Add to dashboard Cite

Most successful object recognition systems rely on binary classification, deciding only if an object is present or not, but not providing information on the actual object location. To perform localization, one can take a sliding window approach, but this strongly increases the computational cost, because the classifier function has to be evaluated over a large set of candidate subwindows.In this paper, we propose a simple yet powerful branchand-bound scheme that allows efficient maximization of a large class of classifier functions over all possible subimages. It converges to a globally optimal solution typically in sublinear time. We show how our method is applicable to different object detection and retrieval scenarios. The achieved speedup allows the use of classifiers for localization that formerly were considered too slow for this task, such as SVMs with a spatial pyramid kernel or nearest neighbor classifiers based on the χ 2 -distance. We demonstrate state-of-the-art performance of the resulting systems on the UIUC Cars dataset, the PASCAL VOC 2006 dataset and in the PASCAL VOC 2007 competition.

show abstract

Section: Pascal Voc 2006 Datasetmentioning

confidence: 78%

Beyond sliding windows: Object localization by efficient subwindow search

Lampert¹,

Blaschko²,

Hofmann

2008

2008 IEEE Conference on Computer Vision and Pattern Recognition

584

458

View full text Add to dashboard Cite

show abstract

“…ESS overall achieves higher overlap with ground truth than any of the sliding window methods. method \ data set cat dog ESS w/ bag-of-visual-words kernel 0.223 0.148 Viitaniemi/Laaksonen [33] 0.179 0.131 Shotton/Winn [32] 0.151 0.118 software provided in the PASCAL VOC challenges: from the precision-recall curves, the average precision (AP) measure is calculated, which is the average of the maximal precision within different intervals of recall, see [32] for details. Table II contains the results, showing that ESS improves over the best results that have been achieved in the VOC 2006 competition or in later publications.…”

Section: A Pascal Voc 2006 Datasetmentioning

confidence: 99%

Efficient Subwindow Search: A Branch and Bound Framework for Object Localization

Lampert

Blaschko

Hofmann

2009

IEEE Trans. Pattern Anal. Mach. Intell.

318

335

View full text Add to dashboard Cite

Abstract-Most successful object recognition systems rely on binary classification, deciding only if an object is present or not, but not providing information on the actual object location. To estimate the object's location one can take a sliding window approach, but this strongly increases the computational cost, because the classifier or similarity function has to be evaluated over a large set of candidate subwindows.In this paper, we propose a simple yet powerful branch and bound scheme that allows efficient maximization of a large class of quality functions over all possible subimages. It converges to a globally optimal solution typically in linear or even sublinear time, in constrast to the quadratic scaling of exhaustive or sliding window search. We show how our method is applicable to different object detection and image retrieval scenarios. The achieved speedup allows the use of classifiers for localization that formerly were considered too slow for this task, such as SVMs with a spatial pyramid kernel or nearest neighbor classifiers based on the χ 2 -distance. We demonstrate state-ofthe-art localization performance of the resulting systems on the UIUC Cars dataset, the PASCAL VOC 2006 dataset and in the PASCAL VOC 2007 competition.

show abstract

“…This approach has also been combined with a discriminatively trained classifier to improve performance [12]. Alternatively, [13] have taken the approach of computing image segments in an unsupervised fashion and cast the localization problem as determining whether each of the segments is an instance of the object. Sliding window classifiers are among the most popular approaches to object localization [1,2,3,4,5,6,7], and the work presented in this paper can broadly be seen to fall into this category.…”

Section: Related Workmentioning

confidence: 99%

Learning to Localize Objects with Structured Output Regression

Blaschko

Lampert

2008

Lecture Notes in Computer Science

244

219

View full text Add to dashboard Cite

Abstract. Sliding window classifiers are among the most successful and widely applied techniques for object localization. However, training is typically done in a way that is not specific to the localization task. First a binary classifier is trained using a sample of positive and negative examples, and this classifier is subsequently applied to multiple regions within test images. We propose instead to treat object localization in a principled way by posing it as a problem of predicting structured data: we model the problem not as binary classification, but as the prediction of the bounding box of objects located in images. The use of a joint-kernel framework allows us to formulate the training procedure as a generalization of an SVM, which can be solved efficiently. We further improve computational efficiency by using a branch-and-bound strategy for localization during both training and testing. Experimental evaluation on the PASCAL VOC and TU Darmstadt datasets show that the structured training procedure improves performance over binary training as well as the best previously published scores.

show abstract

Techniques for Still Image Scene Classification and Object Detection

Cited by 23 publications

References 3 publications

Beyond sliding windows: Object localization by efficient subwindow search

Beyond sliding windows: Object localization by efficient subwindow search

Efficient Subwindow Search: A Branch and Bound Framework for Object Localization

Learning to Localize Objects with Structured Output Regression

Contact Info

Product

Resources

About