Cascade object detection with deformable part models

By Pedro F. Felzenszwalb, Ross B. Girshick, David A. McAllester. 2010

1,006 citations recorded by the source corpus.

Subfields: Artificial Intelligence, 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 Artificial Intelligence

1. Luc Van Gool
2. Trevor Darrell
3. Bernt Schiele
4. Dacheng Tao
5. Li Fei-Fei
6. Xiaogang Wang
7. Jitendra Malik
8. Andrew Zisserman

In Computer Vision and Pattern Recognition

1. Luc Van Gool
2. Xiaogang Wang
3. Dacheng Tao
4. Ming-Hsuan Yang
5. Jitendra Malik
6. Trevor Darrell
7. Bernt Schiele
8. Andrew Zisserman

Citations in the corpus

• Focal Loss for Dense Object Detection (2017, methodology): “ping [17], [18] or hard examplemining [16], [19], [20].”
• Recurrent Models of Visual Attention (2014, background): “…window paradigm, focusing primarily on reducing the number of windows for which the full classifier is evaluated, e.g. via classifier cascades (e.g. [7, 24]), removing image regions from consideration via a branch and bound approach on the classifier output (e.g. [13]), or by proposing candidate…”
• The Open Images Dataset V4 (2018, methodology): “The advent of object detection came in the form classifiers applied densely to windows sliding over the image (e.g. based on boosting (Viola and Jones, 2001a,b) or Deformable Part Models (Felzenszwalb et al., 2010a,b)).”
• Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials (2011, methodology): “For the VOC 2010 dataset we include the response of bounding box object detectors [4] for each object class as 20 additional features.”
• Video In Sentences Out (2012, methodology): “Second, with the star detector (Felzenszwalb et al., 2010b), one can simply decrease the single trained acceptance threshold to yield more detections with no increase in computational complexity.”
• HOGgles: Visualizing Object Detection Features (2013, methodology): “Human accuracy on inversions and state-ofthe-art object detection AP scores from [7] are correlated (a) Human Vision (b) HOG Vision”
• Articulated Human Detection with Flexible Mixtures of Parts (2013, methodology): “Our model requires roughly 1 second to process a typical benchmark image, allowing for the possibility of real-time performance with further speedups (such as cascaded [5] or parallelized implementations).”
• DeepID-Net: multi-stage and deformable deep convolutional neural networks for object detection (2014, neutral): “i-stagetraining Motivation. Multi-stage classifiers have been widely used in object detection and achieved great success. With a cascaded structure, each classifier processes a different subset of data [54, 15, 5, 19, 53]. However, these classifiers are usually trained sequentially without joint optimi”
• G-CNN: An Iterative Grid Based Object Detector (2015, background): “Several extension of DPMs emerged [4, 1] until the remarkable improvements due to the convolutional neural networks was shown by [6].”
• Fast, Accurate Detection of 100,000 Object Classes on a Single Machine (2013, methodology): “At 5 seconds per image, we obtain a speed-up of approximately 20× over the base exhaustive system, which compares well with cascade-based approaches for speeding up detection [7].”
• Deep Regionlets for Object Detection (2017, background): “Deformable Part Model based detection (DPM) [7] further extended the cascaded detectors to more general object categories and proposed the concept of deformable part models to handle the object deformation.”
• Beyond pixels: A comprehensive survey from bottom-up to semantic image segmentation and cosegmentation (2015, background): “The substantial development of image classification [24], object detection [25], superpixel segmentation [3] and 3D scene recovery [26] in the past few years have boosted the research in supervised scene parsing [27], [28], [29], [30], [31], [32].”
• A category-level 3-D object dataset: Putting the Kinect to work (2011, background): “This observation can be leveraged as part of an “objectness” filter or as a thresholding step in a cascaded implementation of this detector for detection speed gain [5, 12].”
• FoveaBox: Beyond Anchor-based Object Detector (2019, methodology): “To add the localization functionality to generic object detection systems, sliding window approaches have been the method of choice for many years [25][7].”
• Latent Hierarchical Model of Temporal Structure for Complex Activity Classification (2014, methodology): “Inspired by the method of cascade object detection in [47], we design a cascade inference algorithm for LHM.”
• Regionlets for Generic Object Detection (2013, methodology): “The exhaust ive search can be accelerated by more efficient search methods as in [5], [28], [4], [6], [26], [23], [29].”
• Joint 3D Object and Layout Inference from a Single RGB-D Image (2015, background): “3D indoor scene understanding is a fundamental problem in computer vison and has recently witnessed great progress enabled by the increasing performance of semantic segmentation and object detection algorithms [6, 10] as well as the availability of RGB-D sensors.”