Roberto Leyva

Video Anomaly Detection With Compact Feature Sets for Online Performance

Over the past decade, video anomaly detection has been explored with remarkable results. However, research on methodologies suitable for online performance is still very limited. In this paper, we present an online framework for video anomaly detection. The key aspect of our framework is a compact set of highly descriptive features, which is extracted from a novel cell structure that helps to define support regions in a coarse-to-fine fashion. Based on the scene’s activity, only a limited number of support regions are processed, thus limiting the size of the feature set. Specifically, we use foreground occupancy and optical flow features. The framework uses an inference mechanism that evaluates the compact feature set via Gaussian Mixture Models, Markov Chains, and Bag-of-Words in order to detect abnormal events. Our framework also considers the joint response of the models in the local spatio-temporal neighborhood to increase detection accuracy. We test our framework on popular existing data sets and on a new data set comprising a wide variety of realistic videos captured by surveillance cameras. This particular data set includes surveillance videos depicting criminal activities, car accidents, and other dangerous situations. Evaluation results show that our framework outperforms other online methods and attains a very competitive detection performance compared with state-of-the-art non-online methods.

The LV dataset: A realistic surveillance video dataset for abnormal event detection

In recent years, designing and testing video anomaly detection methods have focused on synthetic or unrealistic sequences. This has mainly four drawbacks: 1) events are controlled and predictable because they are usually performed by actors; 2) environmental conditions, e.g. camera motion and illumination, are usually ideal thus realistic conditions are not well reflected; 3) events are usually short and repetitive; and 4) the material is captured from scenarios that do not necessarily match the testing scenarios. This leads us to propose a new rich collection of realistic videos captured by surveillance cameras in challenging environmental conditions, the Live Videos (LV) dataset. We explore the performance of a number of state-of-the-art video anomaly detection methods on the LV dataset. Our results confirm the need to design methods that are capable of handling realistic videos captured by surveillance cameras with acceptable processing times. The proposed LV dataset, thus, will facilitate the design and testing of such new methods.

Cross View Gait Recognition Using Joint-Direct Linear Discriminant Analysis

This paper proposes a view-invariant gait recognition framework that employs a unique view invariant model that profits from the dimensionality reduction provided by Direct Linear Discriminant Analysis (DLDA). The framework, which employs gait energy images (GEIs), creates a single joint model that accurately classifies GEIs captured at different angles. Moreover, the proposed framework also helps to reduce the under-sampling problem (USP) that usually appears when the number of training samples is much smaller than the dimension of the feature space. Evaluation experiments compare the proposed framework’s computational complexity and recognition accuracy against those of other view-invariant methods. Results show improvements in both computational complexity and recognition accuracy.

A fast binary pair-based video descriptor for action recognition

Inspired by the binary-based descriptors (e.g. LBP, ALOHA, FREAK, BRISK), we propose the 3D Binary Pair Differences (3DBPD) video descriptor for action recognition. By comparing several spatio-temporal sub-regions around interests points, our descriptor is a feature vector with a dimensionality of up to 30% smaller than that of existing state-of-the-art descriptors. We demonstrate the effectiveness of the 3DBPD descriptor for action recognition with a SVM classifier and a simple Bag Of Video Words (BOV) generated using k-means. The proposed descriptor has very competitive recognition rates compared to other state-of-the-art descriptors, with an outstanding performance in terms of speed. Additionally, the 3DBPD descriptor requires a small codebook compared to those required by existing BOV-based descriptors.

Abnormal event detection in videos using binary features

In this paper we address the problem of online video abnormal event detection. A vast number of methods to automatically detect abnormal events in videos have been recently proposed. However, the majority of these recently proposed methods cannot attain online performance; in other words, they cannot detect events as soon as they occur. Thus there is a lack of methods specifically aimed to detect events in online fashion. In this paper, we propose to incorporate binary features to detect abnormal events in an online manner. This is based on the fact that binary features are well known to require short processing times, compared to double-precision features. The main contribution of this work is then at the feature extraction step. Our experiment results of our binary-based framework show that our proposed binary features help to reduce processing times for anomaly detection, while outperforming other online methods, in terms of detection accuracy.

View-Invariant Gait Recognition Using a Joint-DLDA Framework

In this paper, we propose a new view-invariant framework for gait analysis. The framework profits from the dimensionality reduction advantages of Direct Linear Discriminant Analysis (DLDA) to build a unique view-invariant model. Among these advantages is the capability to tackle the under-sampling problem (USP), which commonly occurs when the number of dimensions of the feature space is much larger than the number of training samples. Our framework employs Gait Energy Images (GEIs) as features to create a single joint model suitable for classification of various angles with high accuracy. Performance evaluations shows the advantages of our framework, in terms of computational time and recognition accuracy, as compared to state-of-the-art view-invariant methods.

Fast Binary-Based Video Descriptors for Action Recognition

Action recognition is one of the top challenges in computer vision. In this paper, we present two binary-based video descriptors with outstanding characteristics in terms of recognition rate, computational times and memory requirements. The descriptors are called Binary Wavelet Differences (BWD) and Binary Dense Trajectories (BDT). Our proposed descriptors are based on the local binary patterns and produce binary vectors with a very low dimensionality. Specifically, we propose to analyze the spatio-temporal support regions of a video sequence to generate binary strings via wavelets patterns. We also propose to encode the motion information obtained from optical flow into a compact binary representation. Our evaluations on the KTH and UCF50 datasets demonstrate that our proposed descriptors achieve very competitive recognition accuracy. Moreover, they are able to attain shorter computational times and smaller memory requirements. Specifically, our proposed descriptors can be calculated up to 20X faster than orientation-based descriptors and require up to 225X less memory. Due to its binary nature, associated calculations in action recognition, e.g. clustering and classification, can be done up to 40X faster than state-of-the-art descriptors. Finally, our descriptors require codebooks with 2X fewer words than those required by other state-of-the-art descriptors.

A view-invariant gait recognition algorithm based on a joint-direct linear discriminant analysis

This paper proposes a view-invariant gait recognition algorithm, which builds a unique view invariant model taking advantage of the dimensionality reduction provided by the Direct Linear Discriminant Analysis (DLDA). Proposed scheme is able to reduce the under-sampling problem (USP) that appears usually when the number of training samples is much smaller than the dimension of the feature space. Proposed approach uses the Gait Energy Images (GEIs) and DLDA to create a view invariant model that is able to determine with high accuracy the identity of the person under analysis independently of incoming angles. Evaluation results show that the proposed scheme provides a recognition performance quite independent of the view angles and higher accuracy compared with other previously proposed gait recognition methods, in terms of computational complexity and recognition accuracy.