Locally-Transferred Fisher Vectors for Texture Classification

  title={Locally-Transferred Fisher Vectors for Texture Classification},
  author={Yang Song and Fan Zhang and Qing Li and Heng Huang and Lauren J. O’Donnell and Weidong (Tom) Cai},
  journal={2017 IEEE International Conference on Computer Vision (ICCV)},
Texture classification has been extensively studied in computer vision. [] Key Method In particular, we design a locally-transferred Fisher vector (LFV) method, which involves a multi-layer neural network model containing locally connected layers to transform the input FV descriptors with filters of locally shared weights. The network is optimized based on the hinge loss of classification, and transferred FV descriptors are then used for image classification. Our results on three challenging texture image…

Figures and Tables from this paper

Deep Texture Recognition via Exploiting Cross-Layer Statistical Self-Similarity‡
A novel feature aggregation module called CLASS (Cross-Layer Aggregation of Statistical Self-similarity) for texture recognition, which encodes both cross-layer dynamics and local SSS of input image, providing additional discrimination over the often-used global average pooling.
Robust Deep Gaussian Descriptor for Texture Recognition
AGaussian descriptor is introduced into B-CNN and a novel robust deep Gaussian descriptor (RDGD) method for texture recognition is proposed, which is superior to its baseline B- CNN and the state-of-the-arts.
Hybrid Featured based Pyramid Structured CNN for Texture Classification
  • Haoran LiuS. KamataYuqi Li
  • Computer Science
    2019 IEEE International Conference on Signal and Image Processing Applications (ICSIPA)
  • 2019
A novel end-to-end structure to make use of hybrid features by a mixture network and improve the classification accuracy, mainly combining Gray Level Co-occurrence Matrix (GLCM) statistical features together with pyramid structured deep convolutional neural networks (Pyramid CNNs) features in a paralleling network structure is proposed.
Texture Classification Using Pair-Wise Difference Pooling-Based Bilinear Convolutional Neural Networks
Since the dimensionality of the BCNN feature vectors is very high, a new yet simple Block-wise PCA (BPCA) method is proposed in order to derive more compact feature vectors.
Multi-layer Feature Fusion and Selection from Convolutional Neural Networks for Texture Classification
This paper aggregates CNN activations from different convolutional layers and encoding them into a single feature vector after applying a pooling operation, and involves a feature selection step that outperforms the state-of-the-art methods with a significant margin.
Using Non-Additive Entropy to Enhance Convolutional Neural Features for Texture Recognition
The use of a local transform that associates each pixel with a measure of local entropy and use such alternative representation as the input to a pretrained convolutional network that performs feature extraction to demonstrate the potential of computing deep learning features over an entropy representation.
Encoding Spatial Distribution of Convolutional Features for Texture Representation
Fractal Encoding is proposed, a feature encoding module grounded by multi-fractal geometry that enables a CNN to encode the regularity on the spatial arrangement of image features, leading to a robust yet discriminative spectrum descriptor.
DenseNet model combined with Haralick’s handcrafted features for texture classification
This paper shows that combining DenseNet model with Haralick’s handcrafted in intermediate layers is a way to overcome the problem of texture representation in CNNs.
Confidence-based Local Feature Selection for Material Classification
This paper proposes to select the most important local features before applying the GAP, and adds a branch in the classification network that predicts the confidence the network should have in each local feature vector.
A Class-Independent Texture-Separation Method Based on a Pixel-Wise Binary Classification
It is shown that the method to segment textures through a binary pixel-wise classification, thereby without the need for a predefined number of textures classes has a good performance on different test sets, can precisely identify borders between texture regions and does not suffer from over-segmentation.


Texture image classification with discriminative neural networks
A discriminative neural network-based feature transformation (NFT) method is designed, with which the CNN-based features are transformed to lower dimensionality descriptors based on an ensemble of neural networks optimized for the classification objective.
Visualizing and Understanding Deep Texture Representations
A systematic evaluation of recent CNN-based texture descriptors for recognition and a technique to visualize pre-images is proposed, providing a means for understanding categorical properties that are captured by these representations.
Deep filter banks for texture recognition and segmentation
This work proposes a new texture descriptor, FV-CNN, obtained by Fisher Vector pooling of a Convolutional Neural Network (CNN) filter bank, which substantially improves the state-of-the-art in texture, material and scene recognition.
Deep FisherNet for Object Classification
The proposed FisherNet combines convolutional neural network training and Fisher Vector encoding in a single end-to-end structure that is learnable using backpropagation and observes a clear advantage of FisherNet over plain CNN and standard FV in terms of both classification accuracy and computational efficiency on the challenging PASCAL VOC object classification task.
Learning and Transferring Mid-level Image Representations Using Convolutional Neural Networks
This work designs a method to reuse layers trained on the ImageNet dataset to compute mid-level image representation for images in the PASCAL VOC dataset, and shows that despite differences in image statistics and tasks in the two datasets, the transferred representation leads to significantly improved results for object and action classification.
Improving the Fisher Kernel for Large-Scale Image Classification
In an evaluation involving hundreds of thousands of training images, it is shown that classifiers learned on Flickr groups perform surprisingly well and that they can complement classifier learned on more carefully annotated datasets.
Deep Fisher Networks for Large-Scale Image Classification
This paper proposes a version of the state-of-the-art Fisher vector image encoding that can be stacked in multiple layers, and significantly improves on standard Fisher vectors, and obtains competitive results with deep convolutional networks at a smaller computational learning cost.
Describing Textures in the Wild
This work identifies a vocabulary of forty-seven texture terms and uses them to describe a large dataset of patterns collected "in the wild", and shows that they both outperform specialized texture descriptors not only on this problem, but also in established material recognition datasets.
Local Features and Kernels for Classification of Texture and Object Categories: An In-Depth Study
This paper presents a large-scale evaluation of an approach that represents images as distributions of features extracted from a sparse set of keypoint locations and learns a Support Vector Machine classifier with kernels based on two effective measures for comparing distributions, the Earth Mover's Distance and the chi-square distance.
Return of the Devil in the Details: Delving Deep into Convolutional Nets
It is shown that the data augmentation techniques commonly applied to CNN-based methods can also be applied to shallow methods, and result in an analogous performance boost, and it is identified that the dimensionality of the CNN output layer can be reduced significantly without having an adverse effect on performance.