Learning a category independent object detection cascade

By Esa Rahtu, Juho Kannala, Matthew B. Blaschko. 2011

185 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Media Technology

How researchers see this paper

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

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

What Makes for Effective Detection Proposals? (2015, methodology): “However, with improved algorithmic efficiency and increased computational power, it is now standard practice to use dense feature extraction [32].”
Self-taught object localization with deep networks (2014, methodology): “Objectness methods [2, 6, 30, 4, 3, 16, 36, 25, 21, 26, 32, 14] aim at generating bounding boxes that yield high recall for a high number of candidates, i.”
How good are detection proposals, really? (2014, methodology): “Search[30] SS 10 ?? ??? ?? Rahtu2011[24] R1 3 · · ? Rand.”
BING: Binarized normed gradients for objectness estimation at 300fps (2014, unknown): “We extensively evaluate our method on the challenging PASCAL VOC2007 [27] and Microsoft COCO [56] [26], Objectness [4], GOP [49], LPO [50], Rahtu [66], RandomPrim [61], Rantalankila [67], and SelectiveSearch [75] 7 using publicly available code.”
Salient Object Detection: A Discriminative Regional Feature Integration Approach (2013, background): “A category independent object detection cascade, which uses superpixel boundary integral, edge distribution and window symmetry to describe objectness, is learnt to rank a number of object window candidates [39].”
Generating Object Segmentation Proposals Using Global and Local Search (2014, methodology): “In addition, there are approaches which try to reduce the number of processed windows by sampling the space of bounding boxes non-uniformly and by discarding nonobject windows in a category-independent manner [2] [14].”
Objectness to improve the bag of visual words model (2014, methodology): “In the second experiment, the SVM classifier based on the one versus all scheme is used for the multi-class task.”
Regionlets for Generic Object Detection (2013, unknown): “Performing object detection with object proposals has been attracting increasing attentions [ 31], [32], [33], [34], [35], [8].”
Salient object detection: A survey (2014, background): “hat some generic characteristics, such as boundary closure, high local contrast, and sometimes visual saliency, are shared by objects of any categories. Boundary edge distribution and window symmetry [177], binarized normed gradients (BING) feature resized to a ﬁxed size (e.g., 8 8) [57], and the nu”
Online Video SEEDS for Temporal Window Objectness (2013, unknown): “In this way, we are able to compare it to s-o-a objectness measures [1, 11, 6, 14].”
Fast, Accurate Detection of 100,000 Object Classes on a Single Machine (2013, methodology): “Alternative approaches to reducing L advocate the use of interest points in the form of jumping windows [19], salience operators like objectness [1, 14], and segmentations [18, 9] as cues for generating object-like window proposals, thus pruning out most of the background regions that a naive approa”
Oriented Object Proposals (2015, methodology): “Zhang et al. [31] and Rahtu et al. [24] extend this idea to use cascaded ranking SVM and structured output ranking, respectively, to combine different cues.”
Deep-learning-based object-level contour detection with CCG and CRF optimization (2017, unknown): “80 Rahtu [18] 0.”
Prime Object Proposals with Randomized Prim's Algorithm (2013, methodology): “…VOC2012 [11] and SUN2012 [29] benchmark data sets and show the superiority of RP in speed and performance compared to state-ofthe-art methods [2] [21] [27], especially in difficult scenarios such as IoU ≥ 0.6: 1000 object proposals are obtained in less than 0.7s while detecting 74% of the…”
SCALPEL: Segmentation Cascades with Localized Priors and Efficient Learning (2013, methodology): “First, we sample bounding boxes from three different publicly available bounding box generation methods: purely bottom-up boxes from the segmentation hierarchy of gPb-owt-ucm [2], categoryindependent boxes from [16], and purely top-down classspecific boxes from [9].”
Relief Impression Image Detection : Unsupervised Extracting Objects Directly From Feature Arrangements of Deep CNN (2016, methodology): “Rahtu [27] use many low level clues to generate boxes, and a scoring method very similiar to Objectness.”
Object proposal by multi-branch hierarchical segmentation (2015, methodology): “Methods such as [1, 22, 9, 7, 28] use features like saliency, image gradient and contour information to measure the objectness of a given sliding window.”
Action Localization with Tubelets from Motion (2014, background): “Rather than reducing the number of sliding windows, category-independent object proposals have been proposed to aid object localization [1, 7, 17, 20].”