Object Detection Using Strongly-Supervised Deformable Part Models

By Hossein Azizpour, Ivan Laptev. 2012

229 citations recorded by the source corpus.

Subfields: 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 Computer Vision and Pattern Recognition

Citations in the corpus

Growing Interpretable Part Graphs on ConvNets via Multi-Shot Learning (2016, methodology): “semanticizing CNN patterns: Part localization/detection is an important task in computer vision (Azizpour and Laptev 2012; Simon, Rodner, and Denzler 2014; Li et al. 2013; Chen et al. 2014).”
The application of two-level attention models in deep convolutional neural network for fine-grained image classification (2014, methodology): “t detectors in unsupervised or latent manner [21, 5]. To further improve the performance, part level annotation is also used in training phase to learn strongly-supervised deformable part-based model [1, 25] or directly used to ﬁnetune pre-trained CNN [4]. Our work is also closely related to recentl”
Viewpoints and keypoints (2014, methodology): “Though similar annotations or fitted CAD models have been successfully used to train better object detection systems [3] as well as for simultaneous object detection and viewpoint estimation [30], the task of keypoint prediction has largely been unaddressed for generic object categories.”
HyperNet: Towards Accurate Region Proposal Generation and Joint Object Detection (2016, methodology): “The DPM [10] and its variants [7][3] have been the dominating methods for years.”
Interpreting CNN knowledge via an Explanatory Graph (2017, methodology): “SS-DPMPart) [2] and PL-DPM-Part [13] and a graphical model for part localization, namely Part-Graph [4].”
Histograms of Sparse Codes for Object Detection (2013, background): “Most approaches have explored model structure, either through non-parametric mixtures or exemplars [21, 10], compositional grammar structure [15], supervised correspondences [5, 2], and low-dimensional projections [29, 25].”
Object-Part Attention Model for Fine-Grained Image Classification (2017, methodology): “It is noted that the CNN used in FOAF is pre-trained not only on ImageNet 1K dataset [32] but also on the dataset of PASCAL VOC [61], while our approach does not use the external dataset like PASCAL VOC.”
Bi-box Regression for Pedestrian Detection and Occlusion Estimation (2018, methodology): “A deformable part model [13] and its variants [15, 2, 42] can also be used for handling occlusions.”
DeePM: A Deep Part-Based Model for Object Detection and Semantic Part Localization (2015, methodology): “This enables ﬂexible part type sharing among different object conﬁgurations, which is different from the tied object-part types in previous DPM models (Felzenszwalb et al., 2010; Azizpour & Laptev, 2012; Chen et al., 2014) on object detection.”
Mining Object Parts from CNNs via Active Question-Answering (2017, methodology): “The fifth and sixth baselines were the strongly supervised DPM (SS-DPM-Part) [2] and the technique in [18] (PL-DPM-Part), respectively.”
Selective Convolutional Descriptor Aggregation for Fine-Grained Image Retrieval (2016, background): “, head and torso) reported in some previous part localization based fine-grained classification algorithms [4, 5,29].”
Detect What You Can: Detecting and Representing Objects Using Holistic Models and Body Parts (2014, methodology): “, [9] or [1], which allows missing body parts) or just a set of templates for parts (e.”
Capturing Long-Tail Distributions of Object Subcategories (2014, background): “Subcategory discovery: Estimating subcategories is surprisingly hard; clustering based on keypoints [16, 5] and appearance [11, 3] have provided only modest performance increases [34].”
Weakly Supervised Learning of Part Selection Model with Spatial Constraints for Fine-Grained Image Classification (2017, methodology): “Bounding box and Part annotations are directly used in training phase to learn a strongly supervised deformable part-based model (Zhang et al. 2013; Azizpour and Laptev 2012) or directly used to fine-tune the pre-trained Convolutional Neural Net (CNN) (Branson et al.”
Learning Deep Representation with Large-Scale Attributes (2015, methodology): “In existing datasets, only the small datasets in [1, 8] have labels on object parts.”
FoveaBox: Beyound Anchor-Based Object Detection (2019, methodology): “The DPM [22] and its variants [23] have been the dominating methods for years.”
Multi-scale Location-Aware Kernel Representation for Object Detection (2018, methodology): “R-CNN greatly improves traditional DPM [9] and its variants [2], where handcrafted features are employed.”
Learning Discriminative Part Detectors for Image Classification and Cosegmentation (2013, unknown): “The deformable part model (DPM) [15] represents an object by a set of deformable parts learned from object bounding boxes.”