Deep Fisher Networks for Large-Scale Image Classification

By Karen Simonyan, Andrea Vedaldi, Andrew Zisserman. 2013

212 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

ImageNet Large Scale Visual Recognition Challenge (2014, background): “The Third Research Institute of the Ministry of Public Security, P.R. China Jie Shao, Xiaoteng Zhang, Yanfeng Shang, Wenfei Wang, Lin Mei, Chuanping Hu UCLA - - 9.8 University of California Los Angeles Yukun Zhu, Jun Zhu, Alan Yuille UIUC - - 1.0 University of Illinois at Urbana-Champaign Thomas Pai”
Two-Stream Convolutional Networks for Action Recognition in Videos (2014, methodology): “Recent improvements of trajectory-based hand-crafted representations include compensation of global (camera) motion [10, 16, 26], and the use of the Fisher vector encoding [22] (in [26]) or its deeper variant [23] (in [21]).”
Compressing Deep Convolutional Networks using Vector Quantization (2014, background): “Deep convolutional neural network has achieved great successes in image classification (Krizhevsky et al., 2012; Jia, 2013; Donahue et al., 2013; Simonyan et al., 2013; Sermanet et al., 2013; Zeiler & Fergus, 2013), object detection (Girshick et al., 2014), and image retrieval (Razavian et al.,…”
Deep Learning: Methods and Applications (2014, methodology): “The VGG system is described in [329] and uses a combination of the deep Fisher vector network and the deep-CNN.”
Understanding Image Representations by Measuring Their Equivariance and Equivalence (2014, background): “(2012) and related state-of-the-art architectures (He et al. 2016; Simonyan et al. 2013), mainly for translation invariance which can be extended for different groups as well (Cohen andWelling 2016; Dieleman et al.”
Towards Open World Recognition (2014, background): “Though more advanced features are available [30, 17, 37], extensive evaluation across features is beyond the scope of this work 2.”
Fisher Vectors Derived from Hybrid Gaussian-Laplacian Mixture Models for Image Annotation (2014, methodology): “Simonyan et al. [39] were motivated by the recent success of Convolutional Neural Networks (CNN) [21] and proposed a version of the state-of-the-art Fisher Vector image encoding that can be stacked in multiple layers.”
Multi-scale Orderless Pooling of Deep Convolutional Activation Features (2014, methodology): “This has spurred a flurry of activity on further improving CNN architectures and training algorithms [3,4,5,6,7], as well as on using CNN features as a universal representation for recognition.”
Self-taught object localization with deep networks (2014, background): “In recent years deep networks have gained great popularity due to their outstanding performance on different tasks, including large-scale image categorization [20, 25], face verification [29], and video recognition [18].”
Learnable pooling with Context Gating for video classification (2017, methodology): “As opposed to [38], [46], these parameters are not related to a Gaussian Mixture Model but instead are trained in a discriminative manner.”
Action Recognition with Stacked Fisher Vectors (2014, background): “Recently, improvement has also been observed in shallow but still hierarchically layered models based on traditional encoding methods for object classification [28,25].”
Supervised mid-level features for word image representation (2014, methodology): “a deep Fisher network for image recognition [35], also used very recently for action recognition [23].”
Action Recognition by Learning Deep Multi-Granular Spatio-Temporal Video Representation (2016, methodology): “The further improvements are achieved by the compensation of camera motion [6], and the use of advanced feature encoding methods such as Bag-of-Words (BoW) [15] and Fisher Vectors [19, 22].”
Deep learning for visual understanding: A review (2016, background): “Besides that, deep CNNs and Fisher Vectors (FV) are complementary [36] and can also be combined to signi ﬁ cantly improve the accuracy of image classi ﬁ cation.”
Learning Beyond Human Expertise with Generative Models for Dental Restorations (2018, unknown): “For example, the deep Fisher network [42] incorporates Fisher vector and orderless pooling [43] combines with Vector of Locally Aggregated Descriptors.”
Hierarchical Coding Vectors for Scene Level Land-Use Classification (2016, methodology): “On the two databases, the HCV achieved better performance than the traditional FV, which has shown great success in computer vision [7,20,24,25,30].”
Hierarchical Gaussian Descriptor for Person Re-identification (2016, background): “Although several descriptors are proposed by focusing on CNN like hierarchy [8, 39], they require learning processes for feature extraction in each hierarchy.”
3DmFV: Three-Dimensional Point Cloud Classification in Real-Time Using Convolutional Neural Networks (2018, background): “A network composed of Fisher layers was suggested in [38].”