Aggregating local descriptors into a compact image representation

By Herv'e J'egou, Matthijs Douze, Cordelia Schmid, Patrick Pérez. 2010

2,881 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

Going deeper with convolutions (2014, methodology): “This enables us to easily use a multi-index hash table [4] to index the centers so that the nearest center lookup becomes extremely efficient.”
Aggregating Local Image Descriptors into Compact Codes (2012, methodology): “The efﬁciency problem was partially addressed by the min-Hash approach [13], [14] and the pre-ﬁltering method of [15].”
Group Normalization (2018, methodology): “Higher-level features such as VLAD [29] and Fisher Vectors (FV) [44] are also group-wise features where a group can be thought of as the sub-vector computed with respect to a cluster.”
Describing Textures in the Wild (2013, methodology): “As we will see next, however, IFV signiﬁcantly outperforms VLAD in other texture datasets.”
Understanding deep image representations by inverting them (2014, background): “Most image understanding and computer vision methods build on image representations such as textons [15], histogram of oriented gradients (SIFT [18] and HOG [4]), bag of visual words [3][25], sparse [35] and local coding [32], super vector coding [37], VLAD [9], Fisher Vectors [21], and, lately, dee”
A Survey of Deep Network Solutions for Learning Control in Robotics: From Reinforcement to Imitation (2016, positive): “s and condition changes, which make it suitable for place recognition. Arandjelovic et al. (2016) combined the features extracted from CNN models with Vector of Locally Aggregated Descriptors (VLAD) (Jegou et al., 2010). They built an end-to-end´ model for place recognition task compared with the fo”
Hierarchical Recurrent Neural Encoder for Video Representation with Application to Captioning (2015, methodology): “[41] propose to utilize Vector of Locally Aggregated Descriptors (VLAD) [13] to aggregate frame level ConvNet for video representation, which is unable to capture temporal structure.”
Bilinear CNN Models for Fine-Grained Visual Recognition (2015, methodology): “The Vector of Locally Aggregated Descriptors (VLAD) descriptor [20] aggregates the first order statistics of the SIFT descriptors.”
Visualizing Deep Convolutional Neural Networks Using Natural Pre-images (2015, background): “2010), VLAD (Jégou et al. 2010), Fisher Vectors (Perronnin and Dance 2006), and, lately, deep neural networks, particularly of the convolutional variety (Krizhevsky et al.”
Compact Bilinear Pooling (2015, methodology): “VLAD [14] and Improved Fisher Vector [26] encoding improved over hard vector quantization by including second order information in the descriptors.”
Fisher Vectors Derived from Hybrid Gaussian-Laplacian Mixture Models for Image Annotation (2014, methodology): “e vector representation is usually passed to a classiﬁer that is trained by a suitable machine learning algorithm, e.g., SVM. This paper focuses on the pooling step. There are many pooling techniques [43,13]. However, the leading pooling technique in recent years is the Fisher Vector [34], which pro”
Context Encoding for Semantic Segmentation (2018, methodology): “Then a visual vocabulary (dictionary) is often learned and the global feature statistics are described by classic encoders such as Bag-of-Words (BoW) [8, 13, 26, 44], VLAD [25] or Fisher Vector [42].”
NetVLAD: CNN Architecture for Weakly Supervised Place Recognition (2015, methodology): “Vector of Locally Aggregated Descriptors (VLAD) [29] is a popular descriptor pooling method for both instance level retrieval [29] and image classification [23].”
Multi-scale Orderless Pooling of Deep Convolutional Activation Features (2014, methodology): “Inspired by previous work on spatial and feature space pooling of local descriptors [14,15,16], we propose a novel and simple pooling”
End-to-End Learning of Deep Visual Representations for Image Retrieval (2016, methodology): “Encoding techniques, such as the Fisher Vector (Perronnin and Dance, 2007; Perronnin et al, 2010) or the VLAD descriptor (Jégou et al, 2010) have been used for example by Perronnin et al (2010); Gordo et al (2012); Jégou and Chum (2012); Radenovic et al (2015).”
Kornia: an Open Source Differentiable Computer Vision Library for PyTorch (2019, methodology): “For ease of use, it supported Windows, Mac OS X, and Linux, and has been a reference e.g for efﬁcient implementations of algorithms such as Fisher Vector [47], VLAD [24], SIFT [33], and MSER [35].”
Learnable pooling with Context Gating for video classification (2017, methodology): “The NetVLAD [23] architecture has been proposed for place recognition to reproduce the VLAD encoding [15], but in a differentiable manner, where the clusters are tuned via backpropagation instead of using k-means clustering.”
Geometric Loss Functions for Camera Pose Regression with Deep Learning (2017, methodology): “Traditionally, this has been modelled as an image retrieval problem [6, 9, 53, 45] enabling the use of efficient and scalable retrieval approaches [35, 38] such as Bag-of-Words (BoW) [47], VLAD [17, 10], and Fisher vectors [18].”