Kernel Pooling for Convolutional Neural Networks

By Yin Cui, Feng Zhou, Jiang Wang, Xiao Liu, Yuanqing Lin, Serge J. Belongie. 2017

322 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

In Computer Vision and Pattern Recognition

Citations in the corpus

Do Better ImageNet Models Transfer Better? (2018, methodology): “Several papers have proposed methods to make better use of CNN features and thus improve the efficacy of transfer learning [43, 11, 42, 23, 73, 63, 13, 41, 54].”
Deep Layer Aggregation (2017, methodology): “For fair comparison, we follow the experimental setup of Cui et al. (2017).”
Large Scale Fine-Grained Categorization and Domain-Specific Transfer Learning (2018, methodology): “Recent advances on Convolutional Neural Networks (CNNs) for visual recognition [33, 48, 51, 20] have fueled remarkable progress on FGVC [36, 11, 69].”
Towards Faster Training of Global Covariance Pooling Networks by Iterative Matrix Square Root Normalization (2017, background): “Kernel pooling [4] approximates Gaussian RBF kernel to a given order through compact explicit feature maps, aiming to characterize higher order feature interactions.”
Hierarchical Bilinear Pooling for Fine-Grained Visual Recognition (2018, methodology): “Still, our model achieves the highest √ 91.3 PA-CNN [13] √ 92.6 FCAN [18] 89.1 RA-CNN [5] 92.5 MA-CNN [37] 92.8 B-CNN [17] 90.6 LRBP [12] 90.9 HIHCA [3] 91.7 Improved B-CNN [16] 92.0 BoostCNN [22] 92.1 KP [4] 92.4 HBP 93 .”
Multi-branch and Multi-scale Attention Learning for Fine-Grained Visual Categorization (2020, methodology): “ent on a wide range of visual tasks. However, bilinear feature dimensions are high, usually orders of magnitude in the hundreds of thousands to millions, which limits its further generalization. [9], [10] are improvements. The second kind of method, localization subnetwork, locates key part regions ”
Improving Fine-Grained Visual Classification using Pairwise Confusion (2017, methodology): “Similar to LSVC tasks, neural networks for FGVC tasks are typically trained with cross-entropy loss [1,7,8,9].”
Learning Deep Bilinear Transformation for Fine-grained Image Representation (2019, unknown): “It can be observed that the proposed DBT can signiﬁcantly outperform Compact bilinear [14] and Kernel Pooling [37] in all three datasets.”
Kervolutional Neural Networks (2019, methodology): “s of the complex and hypercomplex cells inside the striate cortex. It was also demonstrated that higher order non-linear feature maps are able to make subsequent linear classiﬁers more discriminative [37, 1, 9]. However, the non-linearity that comes from the activation layers, e.g. rectiﬁed linear u”
Learning Rich Part Hierarchies With Progressive Attention Networks for Fine-Grained Image Recognition (2020, unknown): “, part localization [11]–[14] to normalize part features explicitly and high-order feature [2], [15], [16] to obtain more powerful representation.”
Adaptive Triplet Model for Fine-Grained Visual Categorization (2018, background): “One-step models [22]–[29] are often based on different intuitions and thus are much more diversified.”
Compact Generalized Non-local Network (2018, methodology): “We expend the Taylor series with third order and the hyperparameter γ for RBF is set by 1e-4 [3].”
Multi-view Harmonized Bilinear Network for 3D Object Recognition (2018, methodology): “Recently it is used in pooling local convolutional features [20, 7, 5, 17] to take the second-order statistics into consideration, achieving state-of-the-art performance in fine-grained image classification.”
Multi-Attention Multi-Class Constraint for Fine-grained Image Recognition (2018, methodology): “Our baselines, on the other hand, are trained with simple augmentation ( e.g. , mirror and random cropping) and meet the Caﬀe baselines of other works, such as 82.0% in [32] and 78.4% in [54].”
HoMM: Higher-order Moment Matching for Unsupervised Domain Adaptation (2019, methodology): “tistics The statistics higher than ﬁrstorder has been success fully used in many classical and deep learning methods [9, 19, 30, 22, 12]. Especially in the ﬁeld of ﬁne-grained image/video recognition [24, 8], secondorder statistics such as Covariance and Gaussian descriptors, have demonstrated bette”
Fine-Grained Visual Classification via Progressive Multi-Granularity Training of Jigsaw Patches (2020, unknown): “e-art methods. Method Base Model CUB (%) CAR (%) AIR (%) FT VGG (CVPR18) [31] VGG16 77.8 84.9 84.8 FT ResNet (CVPR18) [31] ResNet50 84.1 91.7 88.5 B-CNN (ICCV15) [22] VGG16 84.1 91.3 84.1 KP (CVPR17) [8] VGG16 86.2 92.4 86.9 RA-CNN (ICCV17) [11] VGG19 85.3 92.5 - MA-CNN (ICCV17) [38] VGG19 86.5 92.8”
Measuring Dataset Granularity (2019, background): “Examples of such modules include non-linear feature interactions [32, 15, 7] and attention mechanisms [22, 53, 14, 56, 46, 34, 55].”