Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories

By Svetlana Lazebnik, Cordelia Schmid, Jean Ponce. 2006

8,884 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

Fast R-CNN (2015, positive): “[10] advocate concatenating outputs from multiple pooling grids ( e.g ., 6 × 6 , 3 × 3 , 2 × 2 , 1 × 1 ), as in spatial pyramid pooling [14].”
Object Detection Networks on Convolutional Feature Maps (2015, methodology): “The Selective Search paper [4] relies on spatial pyramid features [20] on dense SIFT vectors [21] and an additive kernel SVM.”
ImageNet Large Scale Visual Recognition Challenge (2014, methodology): “…by dense sampling and vector quantization of several color SIFT features (van de Sande et al., 2010), pooling with spatial pyramid matching (Lazebnik et al., 2006), and classifying with a histogram intersection kernel SVM (Maji and Malik, 2009) trained on a GPU (van de Sande et al., 2011a).”
Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition (2014, methodology): “Then we still use win = da/ne and str = ba/nc to implement each pyramid pooling level.”
Remote Sensing Image Scene Classification: Benchmark and State of the Art (2017, methodology): “In brief, SPM divides each image into increasingly finer spatial subregions, constructs the BoVW representations for each subregion, and then concatenates them to represent the image.”
SUN database: Large-scale scene recognition from abbey to zoo (2010, methodology): “How many kinds of scenes are there? How can the knowledge about environmental scenes be organized? How do the current state-of-art scene models perform on more realistic and ill-controlled environments, and how does this compare to human performance? To date, computational work on scene and place re”
DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs (2016, methodology): “Instead, motivated by spatial pyramid pooling [19], [20], we propose a computationally efficient scheme of resampling a given feature layer at multiple rates prior to convolution.”
Scattering Networks for Hybrid Representation Learning (2018, methodology): “Two approaches are then used: the spatial localization is handled either by a Spatial Pyramid Pooling [30], which is then fed to a linear SVM, or the spatial variables are directly encoded in the FVs and classified with a stack of four fully connected layers.”
Multiview Transformers for Video Recognition (2022, background): “A central idea, that has remained constant across classical methods based on handcrafted features [9, 14, 38] to CNNs [42, 43, 85] and now transformers [11, 44, 74], has been to analyze input signals at multiple resolutions.”
Representation Learning: A Review and New Perspectives (2012, methodology): “We advocate learning algorithms that are flexible and non-parametric7 but do not rely exclusively on the smoothness assumption.”
Pyramid Scene Parsing Network (2016, methodology): “To make good use of global image-level priors for diverse scene understanding, methods of [18, 27] extracted global context information with traditional features not from deep neural networks.”
Deformable Convolutional Networks (2017, methodology): “Spatial manipulation in RoI pooling Spatial pyramid pooling [34] uses hand crafted pooling regions over scales.”
Learning to See by Moving (2015, methodology): “Using this insight, we pose the problem of correlating visual stimuli with egomotion as the problem of predicting the camera transformation from the consequent pairs of images that the agent receives while it moves.”
Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation (2018, methodology): “In this work, we consider two types of neural networks that use spatial pyramid pooling module [23, 40, 26] or encoder-decoder structure [61, 3] for semantic segmentation, where the former one captures rich contextual information by pooling features at different resolution while the latter one is ab”
Rethinking Atrous Convolution for Semantic Image Segmentation (2017, background): “ASPP is inspired by the success of spatial pyramid pooling [23, 41, 26] which showed that it is effective to resample features at different scales for accurately and efficiently classifying regions of an arbitrary scale.”
Return of the Devil in the Details: Delving Deep into Convolutional Nets (2014, methodology): “In order to capture weak geometrical information, the IFV representation is used in a spatial pyramid [26].”
Analyzing the Performance of Multilayer Neural Networks for Object Recognition (2014, background): “, [12]), or by first finding a set of mid-level patches [10,20] and then encoding images in terms of them.”
ParseNet: Looking Wider to See Better (2015, methodology): “In addition, we also tried to follow the spatial pyramid idea [17] to pool features from increasingly finer subregions and attach them to local features in the sub-regions, however, we did not observe significant improvements.”
Object Detectors Emerge in Deep Scene CNNs (2014, background): “to objects, other featural regularities of scene categories allow for other representations to emerge, such as textures (Renninger & Malik, 2004), GIST (Oliva & Torralba, 2006), bag-of-words (Lazebnik et al., 2006), part-based models (Pandey & Lazebnik, 2011), and ObjectBank (Li et al., ”
Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation (2015, methodology): “We explore background context by applying a network architecture with traditional multi-scale image input [13] and sliding pyramid pooling [26].”