Multiresolution Models for Object Detection

By Dennis Park, Deva Ramanan, Charless C. Fowlkes. 2010

270 citations recorded by the source corpus.

Subfields: Aerospace Engineering, 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 Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Ensemble of exemplar-SVMs for object detection and beyond (2011, background): “Because most detections show good alignment to their associated exemplar, it is possible to transfer any available exemplar meta-data (segmentation, geometric structure, 3D model, etc.) directly onto the detections, which can then be used as part of overall scene understanding.”
DeepID-Net: multi-stage and deformable deep convolutional neural networks for object detection (2014, methodology): “information investigated in previous works includes regions surrounding objects [9, 12, 22], object-scene interaction [13], and the presence, location, orientation and size relationship among objects [3, 57, 58, 11, 41, 22, 49, 13, 61, 12, 59, 40, 10, 45, 51]. In this paper, we utilize the image cla”
Modeling Visual Context is Key to Augmenting Object Detection Datasets (2018, methodology): “Before the deep learning/ImageNet revolution, the previous generation of object detectors such as [3,4,5,6] modeled the interaction between object locations, categories, and context by manual engineering of local descriptors, feature aggregation methods, and by deﬁning structural relationship…”
Scalable multi-class object detection (2011, unknown): “In the future, we are planning to overcome these shortcomings by using texture features and building a multi-resolution representation of patches similar to [23].”
Object detection and segmentation from joint embedding of parts and pixels (2011, background): “Pixels and parts each appear as nodes in a graph whose edges encode both affinity and ordering relationships.”
Ten Years of Pedestrian Detection, What Have We Learned? (2014, methodology): “We encode the optical ﬂow using the same SDt features from ACF+SDt [33] (image diﬀerence between current frame T and coarsely aligned T-4 and T-8).”
Local Decorrelation For Improved Pedestrian Detection (2014, background): “Pedestrian Detection: Recent work in pedestrian detection includes use of deformable part models and their extensions [11] [36] [26], convolutional nets and deep learning [33] [37] [25], and approaches that focus on optimization and learning [20] [18] [34].”
Informed Haar-Like Features Improve Pedestrian Detection (2014, methodology): “MultiResC [21] HOG latent SVM Caltech / 48.”
Real-Time Pedestrian Detection with Deep Network Cascades (2015, methodology): “Some methods focus on improving the base features used [6, 9, 10, 42], whereas others focus on the learning algorithms [13, 25], or other techniques such as incorporating Deformable Parts Models [33, 40] or using context [13, 29, 40, 41].”
Pedestrian detection at 100 frames per second (2012, methodology): “plane has often been used as prior knowledge to improve both speed and quality [9, 14, 17, 3].”
Single-Pedestrian Detection Aided by Multi-pedestrian Detection (2013, methodology): “This experiment shows that the multipedestrian detector provides rich complementary information to current state-of-the-art 1-pedestrian detection approaches even when context [25] or motion [33] is used by these approaches.”
Nonparametric Scene Parsing via Label Transfer (2011, background): “Moreover, the authors in [47] showed that with a database of 80 million images, even simple SSD match can give semantically meaningful parsing for 32 32 images.”
Robust Multi-resolution Pedestrian Detection in Traffic Scenes (2013, methodology): “The most related work is [33], which utilized root and part filters for high resolution pedestrians, while only used the rigid root filter for low resolution pedestrians.”
Finding Tiny Faces (2016, methodology): “Our proposed work makes use of large local context (as opposed to a global contextual descriptor [2, 17]) in a scale-variant way (as opposed to [27]).”
Data-driven exemplar model selection (2014, background): “[28], each captures only one type of variation in the object category - either the pose, the viewpoint or the scale respectively.”
Toward Holistic Scene Understanding: Feedback Enabled Cascaded Classification Models (2012, background): “Park et al. [34] used the ground plane estimation as a contextual information for pedestrian detection.”
Face detection, pose estimation, and landmark localization in the wild (2012, background): “One could define multiresolution variants of our models designed for smaller faces [29], but we leave this as future work.”
Pedestrian Detection: An Evaluation of the State of the Art (2012, background): “As part models seem to be most successful at higher resolutions, Park et al. [84] extended this to a multiresolution model that automatically switches to parts only at sufficiently high resolutions.”
Exploring Weak Stabilization for Motion Feature Extraction (2013, result): “5 we compare with previous work including ‘MultiFtr+Motion’ [28] (which uses motion features) and ‘MultiresC’ [23] (which uses static features trained on the same data as [12]).”