On Detection of Multiple Object Instances Using Hough Transforms

By Olga Barinova, Victor S. Lempitsky, Pushmeet Kohli. 2010

331 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Environmental Engineering, Ocean Engineering

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

In Environmental Engineering

In Ocean Engineering

Citations in the corpus

Learning Non-maximum Suppression (2017, background): “Hough voting establishes correspondences between detections and the image evidence supporting them, which can avoid overusing image content for several detections [15, 1, 14, 34].”
End-to-End People Detection in Crowded Scenes (2015, methodology): “Early work employed a codebook of local features and Hough voting [13, 2], but still required complex tuning and multi-stage pipelines.”
Finely-grained annotated datasets for image-based plant phenotyping (2016, background): “, 2013; He and Gould, 2014) or detection (Barinova et al., 2012) approaches, and others.”
Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation (2019, methodology): “Additionally, our method bears a similarity to Hough-Voting-based methods [4, 39, 21, 5] and recent works by Kendall et al.”
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].”
Unsupervised object discovery and localization in the wild: Part-based matching with bottom-up region proposals (2015, background): “Like all its “probabilistic Hough transform” predecessors we know of [3, 31, 33], however, it lacks a proper probabilistic interpretation.”
Efficient Point Process Inference for Large-Scale Object Detection (2016, methodology): “While [3, 5] detect only dozens of objects from the images, we extract thousands of object instances.”
Diverse M-Best Solutions in Markov Random Fields (2012, background): “Diversity is also a key goal in the context of non-maximal suppression for object detection [2, 3].”
DeepID-Net: Deformable deep convolutional neural networks for object detection (2014, background): “The context information investigated in literature includes regions surrounding objects [5, 8, 13], object-scene interaction [9, 20], and the presence, location, orientation and size relationship among objects [2, 48, 49, 7, 31, 13, 40, 9, 53, 8, 50, 30, 6, 35, 45].”
Joint Deep Learning for Pedestrian Detection (2013, methodology): “It has attracted a great deal of research interest [2, 5, 12, 47, 8].”
Non-maximum Suppression for Object Detection by Passing Messages Between Windows (2014, background): “For instance, with Hough Forests [20–22], patches vote for the location of the object center.”
DeepID-Net: Object Detection with Deformable Part Based Convolutional Neural Networks (2017, unknown): “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].”
Hough Forests for Object Detection, Tracking, and Action Recognition (2011, methodology): “In a separate direction, [ 44 ] proposes a non-maxima suppression scheme for Hough-based detection that copes better with multiple occluding instances.”
Relating Things and Stuff via ObjectProperty Interactions (2014, background): “(6) show that this potential is an Occam razor or MDL prior, similar to [6], [13] so that the model is penalized by g l j Á jX j j when y j ¼ 1.”
Density-aware person detection and tracking in crowds (2011, background): “For example, [29] uses scene priming to constrain locations of objects in the image, [3, 30] seek detections that explain a set of detected parts, whereas [8, 25] model spatial and co-occurence interactions between objects in the scene while [31] attempts to explore relations between hu-”
Learning People Detectors for Tracking in Crowded Scenes (2013, background): “Approaches that operate on monocular images typically handle occlusions by carefully separating the evidence coming from the occluder and the occluded objects, either by reasoning on an image segmentation [14], or by iteratively discarding image evidence corresponding to foreground objects [4, 18].”
A Convnet for Non-maximum Suppression (2015, background): “Linking detections to pixels The Hough voting framework enables more sophisticated reasoning amongst conflicting detections by linking the detections to local image evidence (Leibe et al., 2008; Barinova et al., 2012; Wohlhart et al., 2012).”
Detecting Partially Occluded Objects with an Implicit Shape Model Random Field (2012, methodology): “This was also pointed out in [11], where a probabilistic formulation of the object detection task based on the generalized Hough transformation was presented.”
Unconstrained Salient Object Detection via Proposal Subset Optimization (2016, methodology): “In [5], an improved NMS method is proposed for Hough transform based object detectors.”