Fisher Kernel Framework Research Articles

Tangent Fisher Vector on Matrix Manifolds for Action Recognition.

In this paper, we address the problem of representing and recognizing human actions from videos on matrix manifolds. For this purpose, we propose a new vector representation method, named tangent Fisher vector, to describe video sequences in the Fisher kernel framework. We first extract dense curved spatio-temporal cuboids from each video sequence. Compared with the traditional 'straight cuboids', the dense curved spatio-temporal cuboids contain much more local motion information. Each cuboid is then described using a linear dynamical system (LDS) to simultaneously capture the local appearance and dynamics. Furthermore, a simple yet efficient algorithm is proposed to learn the LDS parameters and approximate the observability matrix at the same time. Each video sequence is thus represented by a set of LDSs. Considering that each LDS can be viewed as a point in a Grassmann manifold, we propose to learn an intrinsic GMM on the manifold to cluster the LDS points. Finally a tangent Fisher vector is computed by first accumulating all the tangent vectors in each Gaussian component, and then concatenating the normalized results across all the Gaussian components. A kernel is defined to measure the similarity between tangent Fisher vectors for classification and recognition of a video sequence. This approach is evaluated on the state-of-the-art human action benchmark datasets. The recognition performance is competitive when compared with current state-of-the-art results.

Correlated Topic Vector for Scene Classification.

Scene images usually involve semantic correlations, particularly when considering large-scale image data sets. This paper proposes a novel generative image representation, correlated topic vector, to model such semantic correlations. Oriented from the correlated topic model, correlated topic vector intends to naturally utilize the correlations among topics, which are seldom considered in the conventional feature encoding, e.g., Fisher vector, but do exist in scene images. It is expected that the involvement of correlations can increase the discriminative capability of the learned generative model and consequently improve the recognition accuracy. Incorporated with the Fisher kernel method, correlated topic vector inherits the advantages of Fisher vector. The contributions to the topics of visual words have been further employed by incorporating the Fisher kernel framework to indicate the differences among scenes. Combined with the deep convolutional neural network (CNN) features and Gibbs sampling solution, correlated topic vector shows great potential when processing large-scale and complex scene image data sets. Experiments on two scene image data sets demonstrate that correlated topic vector improves significantly the deep CNN features, and outperforms existing Fisher kernel-based features.

Retrieval of Brain Tumors by Adaptive Spatial Pooling and Fisher Vector Representation.

Content-based image retrieval (CBIR) techniques have currently gained increasing popularity in the medical field because they can use numerous and valuable archived images to support clinical decisions. In this paper, we concentrate on developing a CBIR system for retrieving brain tumors in T1-weighted contrast-enhanced MRI images. Specifically, when the user roughly outlines the tumor region of a query image, brain tumor images in the database of the same pathological type are expected to be returned. We propose a novel feature extraction framework to improve the retrieval performance. The proposed framework consists of three steps. First, we augment the tumor region and use the augmented tumor region as the region of interest to incorporate informative contextual information. Second, the augmented tumor region is split into subregions by an adaptive spatial division method based on intensity orders; within each subregion, we extract raw image patches as local features. Third, we apply the Fisher kernel framework to aggregate the local features of each subregion into a respective single vector representation and concatenate these per-subregion vector representations to obtain an image-level signature. After feature extraction, a closed-form metric learning algorithm is applied to measure the similarity between the query image and database images. Extensive experiments are conducted on a large dataset of 3604 images with three types of brain tumors, namely, meningiomas, gliomas, and pituitary tumors. The mean average precision can reach 94.68%. Experimental results demonstrate the power of the proposed algorithm against some related state-of-the-art methods on the same dataset.

Discourse Relations Detection via a Mixed Generative-Discriminative Framework

Word embeddings, which can better capture the fine-grained semantics of words, have proven to be useful for a variety of natural language processing tasks. However, because discourse structures describe the relationships between segments of discourse, word embeddings cannot be directly integrated to perform the task. In this paper, we introduce a mixed generative-discriminative framework, in which we use vector offsets between embeddings of words to represent the semantic relations between text segments and Fisher kernel framework to convert a variable number of vector offsets into a fixed length vector. In order to incorporate the weights of these offsets into the vector, we also propose the Weighted Fisher Vector. Experimental results on two different datasets show that the proposed method without using manually designed features can achieve better performance on recognizing the discourse level relations in most cases.

A Comparative Study of Sampling Analysis in the Scene Classification of Optical High-Spatial Resolution Remote Sensing Imagery

Scene classification, which consists of assigning images with semantic labels by exploiting the local spatial arrangements and structural patterns inside tiled regions, is a key problem in the automatic interpretation of optical high-spatial resolution remote sensing imagery. Many state-of-the-art methods, e.g., the bag-of-visual-words model and its variants, the topic models and unsupervised feature learning-based approaches, share similar procedures: patch sampling, feature learning and classification. Patch sampling is the first and a key procedure, and it has a considerable influence on the results. In the literature, many different sampling strategies have been used, e.g., random sampling and saliency-based sampling. However, the sampling strategy that is most suitable for the scene classification of optical high-spatial resolution remote sensing images remains unclear. In this paper, we comparatively study the effects of different sampling strategies under the scenario of scene classification of optical high-spatial resolution remote sensing images. We divide the existing sampling methods into two types: random sampling and saliency-based sampling. Here, we consider the commonly-used grid sampling to be a specific type of random sampling method, and the saliency-based sampling consists of keypoint-based sampling and salient region-based sampling. To compare their performances, we rely on a standard bag-of-visual-words model to learn the global features for testing because of its simplicity, robustness and efficiency. In addition, we conduct experiments using a Fisher kernel framework to validate our conclusions. The experimental results obtained on two commonly-used datasets using different feature learning methods show that random sampling can provide comparable and even better performance than all of the saliency-based strategies.

VoD: A novel image representation for head yaw estimation

Building on the recent advances in the Fisher kernel framework for image classification, this paper proposes a novel image representation for head yaw estimation. Specifically, for each pixel of the image, a concise 9-dimensional local descriptor is computed consisting of the pixel coordinates, intensity, the first and second order derivatives, as well as the magnitude and orientation of the gradient. These local descriptors are encoded by Fisher vectors before being pooled to produce a global representation of the image. The proposed image representation is effective to head yaw estimation, and can be further improved by metric learning. A series of head yaw estimation experiments have been conducted on five datasets, and the results show that the new image representation improves the current state-of-the-art for head yaw estimation.

Analysis and Predictive Modeling of Body Language Behavior in Dyadic Interactions from Multimodal Interlocutor Cues

During dyadic interactions, participants adjust their behavior and give feedback continuously in response to the behavior of their interlocutors and the interaction context. In this paper, we study how a participant in a dyadic interaction adapts his/her body language to the behavior of the interlocutor, given the interaction goals and context. We apply a variety of psychology-inspired body language features to describe body motion and posture. We first examine the coordination between the dyad's behavior for two interaction stances: friendly and conflictive. The analysis empirically reveals the dyad's behavior coordination, and helps identify informative interlocutor features with respect to the participant's target body language features. The coordination patterns between the dyad's behavior are found to depend on the interaction stances assumed. We apply a Gaussian-Mixture-Model-based (GMM) statistical mapping in combination with a Fisher kernel framework for automatically predicting the body language of an interacting participant from the speech and gesture behavior of an interlocutor. The experimental results show that the Fisher kernel-based approach outperforms methods using only the GMM-based mapping, and using the support vector regression, in terms of correlation coefficient and RMSE. These results suggest a significant level of predictability of body language behavior from interlocutor cues.

Image Classification with the Fisher Vector: Theory and Practice

A standard approach to describe an image for classification and retrieval purposes is to extract a set of local patch descriptors, encode them into a high dimensional vector and pool them into an image-level signature. The most common patch encoding strategy consists in quantizing the local descriptors into a finite set of prototypical elements. This leads to the popular Bag-of-Visual words representation. In this work, we propose to use the Fisher Kernel framework as an alternative patch encoding strategy: we describe patches by their deviation from an “universal” generative Gaussian mixture model. This representation, which we call Fisher vector has many advantages: it is efficient to compute, it leads to excellent results even with efficient linear classifiers, and it can be compressed with a minimal loss of accuracy using product quantization. We report experimental results on five standard datasets—PASCAL VOC 2007, Caltech 256, SUN 397, ILSVRC 2010 and ImageNet10K—with up to 9M images and 10K classes, showing that the FV framework is a state-of-the-art patch encoding technique.