Human action recognition is an increasingly important research topic in the fields of video sensing, analysis and understanding. Caused by unconstrained sensing conditions, there exist large intra-class variations and inter-class ambiguities in realistic videos, which hinder the improvement of recognition performance for recent vision-based action recognition systems. In this paper, we propose ...

This paper proposes a novel approach to recognize object and scene categories in depth images. We introduce a Bag of Words (BoW) representation in 3D, the Selective 3D Spatial Pyramid Matching Kernel (3DSPMK). It starts quantizing 3D local descriptors, computed from point clouds, to build a vocabulary of 3D visual words. This codebook is used to build the 3DSPMK, which starts partitioning a wor...

In the recent digital age, support vector machines (SVMs) that use a spatial pyramid matching (SPM) kernel have been around the globe for image classification. Although this is popular, there exists many problems in its use; for examples, nonlinear SVMs have a high complexity in training and testing. Applying the algorithms to big datasets, which holds many images, greater than a thousand is a ...

This working note describes the method of the NUDT team for scene classification and object recognition in the ImageCLEF 2014 Robot Vision Challenge. The method is composed of two steps: 1. spatial pyramid match (SPM) and a Pyramid of HOG (Histograms of Oriented Gradient) are incorporated to represent an indoor place image. 2. a multiclass SVM (Support Vector Machine) is utilized to classify an...

Histogram pyramid representations computed from a vocabulary tree of visual words have proven valuable for a range of image indexing and recognition tasks; however, they have only used a single, fixed partition of feature space. We present a new efficient algorithm to incrementally compute set-of-trees (forest) vocabulary representations, and show they improve recognition and indexing performan...

Real-world image classification, which aims to determine the semantic class of un-labeled images, is a challenging task. In this paper, we focus on two challenges of image classification and propose a method to address both of them simultaneously. The first challenge is that representing images by heterogeneous features, such as color, shape and texture, helps to provide better classification a...

It is often useful to represent a single example by a set of the local features that comprise it. However, this representation poses a challenge to many conventional learning techniques, since sets may vary in cardinality and the elements are unordered. To compare sets of features, researchers often resort to solving for the least-cost correspondences, but this is computationally expensive and ...

In this work, we present a deep convolutional pyramid person matching network (PPMN) with specially designed Pyramid Matching Module to address the problem of person reidentification. The architecture takes a pair of RGB images as input, and outputs a similiarity value indicating whether the two input images represent the same person or not. Based on deep convolutional neural networks, our appr...

The workflow of extracting features from images using convolutional neural networks (CNN) and generating captions with recurrent neural networks (RNN) has become a de-facto standard for image captioning task. However, since CNN features are originally designed for classification task, it is mostly concerned with the main conspicuous element of the image, and often fails to correctly convey info...