Intelligent video surveillance systems are rapidly being introduced to public places. The adoption of computer vision and machine learning techniques enables various applications for collected video features; one of the major is safety monitoring. The efficacy of violent event detection is measured by the efficiency and accuracy of violent event detection. In this paper, we present a novel architecture for violence detection from video surveillance cameras. Our proposed model is a spatial feature extracting a U-Net-like network that uses MobileNet V2 as an encoder followed by LSTM for temporal feature extraction and classification. The proposed model is computationally light and still achieves good results—experiments showed that an average accuracy is 0.82 ± 2% and average precision is 0.81 ± 3% using a complex real-world security camera footage dataset based on RWF-2000.

/pdf/efficient-violence-detection-in-surveillance-qqq48mcl.pdf

Efficient Violence Detection in Surveillance

Introducing efficient automatic violence detection in video surveillance or audiovisual content monitoring systems would greatly facilitate the work of closed-circuit television (CCTV) operators, rating agencies or those in charge of monitoring social network content. In this paper we present a new deep learning architecture, using an adapted version of DenseNet for three dimensions, a multi-head self-attention layer and a bidirectional convolutional long short-term memory (LSTM) module, that allows encoding relevant spatio-temporal features, to determine whether a video is violent or not. Furthermore, an ablation study of the input frames, comparing dense optical flow and adjacent frames subtraction and the influence of the attention layer is carried out, showing that the combination of optical flow and the attention mechanism improves results up to 4.4%. The conducted experiments using four of the most widely used datasets for this problem, matching or exceeding in some cases the results of the state of the art, reducing the number of network parameters needed (4.5 millions), and increasing its efficiency in test accuracy (from 95.6% on the most complex dataset to 100% on the simplest one) and inference time (less than 0.3 s for the longest clips). Finally, to check if the generated model is able to generalize violence, a cross-dataset analysis is performed, which shows the complexity of this approach: using three datasets to train and testing on the remaining one the accuracy drops in the worst case to 70.08% and in the best case to 81.51%, which points to future work oriented towards anomaly detection in new datasets.

ViolenceNet: Dense Multi-Head Self-Attention with Bidirectional Convolutional LSTM for Detecting Violence

There are many problems in existing reconstruction-based super-resolution algorithms, such as the lack of texture-feature representation and of high-frequency details. Multi-scale detail enhancement can produce more texture information and high-frequency information. Therefore, super-resolution reconstruction of remote-sensing images based on adaptive multi-scale detail enhancement (AMDE-SR) is proposed in this paper. First, the information entropy of each remote-sensing image is calculated, and the image with the maximum entropy value is regarded as the reference image. Subsequently, spatio-temporal remote-sensing images are processed using phase normalization, which is to reduce the time phase difference of image data and enhance the complementarity of information. The multi-scale image information is then decomposed using the L0 gradient minimization model, and the non-redundant information is processed by difference calculation and expanding non-redundant layers and the redundant layer by the iterative back-projection (IBP) technique. The different-scale non-redundant information is adaptive-weighted and fused using cross-entropy. Finally, a nonlinear texture-detail-enhancement function is built to improve the scope of small details, and the peak signal-to-noise ratio (PSNR) is used as an iterative constraint. Ultimately, high-resolution remote-sensing images with abundant texture information are obtained by iterative optimization. Real results show an average gain in entropy of up to 0.42 dB for an up-scaling of 2 and a significant promotion gain in enhancement measure evaluation for an up-scaling of 2. The experimental results show that the performance of the AMED-SR method is better than existing super-resolution reconstruction methods in terms of visual and accuracy improvements.

Spatio-Temporal Super-Resolution Reconstruction of Remote-Sensing Images Based on Adaptive Multi-Scale Detail Enhancement.

An image detail enhancement algorithm is proposed based on in-place residual homogeneity (IP). Residual homogeneity is a physical law, which mainly explains the texture similarity between the same image residual at slightly different resolutions. As we all know, a single image can be divided into a base layer and a detail layer, and the effective estimation of the detail layer is the key in a detail enhancement algorithm. In the experiment, we find that the residual layer of an image obtained by bilinear interpolation is closely related to its detail layer, hence it can be used as the initial estimation of the detail layer, then residual homogeneity is applied to update the residual layer until the accurate detail layer is acquired. In the process of updating residuals, a searching method called fast in-place searching (FIPS) is used. FIPS only takes advantage of the residual homogeneity within the in-place region, which accelerates the project about 93%. Different from the local-based and global-based methods, our IP gets the detail layer directly and amplifies it. It has many good properties, such as being fast, edge-aware, robust, and parameter-free. Good performance has been demonstrated on several widely used datasets by both subjective and objective evaluations.

Learning in-place residual homogeneity for single image detail enhancement

Single image detail enhancement via metropolis theorem

A novel framework based on local spatio-temporal features and a Bag-of-Words (BoW) model is proposed for violence detection. The framework utilizes Dense Trajectories (DT) and MPEG flow video descriptor (MF) as feature descriptors and employs Fisher Vector (FV) in feature coding. DT and MF algorithms are more descriptive and robust, because they are combinations of various feature descriptors, which describe trajectory shape, appearance, motion and motion boundary, respectively. FV is applied to transform low level features to high level features. FV method preserves much information, because not only the affiliations of descriptors are found in the codebook, but also the first and second order statistics are used to represent videos. Some tricks, that PCA, K-means++ and codebook size, are used to improve the final performance of video classification. In comprehensive consideration of accuracy, speed and application scenarios, the proposed method for violence detection is analysed. Experimental results show that the proposed approach outperforms the state-of-the-art approaches for violence detection in both crowd scenes and non-crowd scenes.

/pdf/violence-detection-based-on-spatio-temporal-feature-and-20ddi4qheb.pdf

Violence Detection Based on Spatio-Temporal Feature and Fisher Vector

In this paper, a low-complexity yet powerful method based on scalable motion analysis (SMA) is proposed. SMA mainly aims at de-noising, especially for surveillance videos in bad illumination. When we do research on surveillance videos, finding that stationary objects’ noise can be removed by averaging adjacent frames, but moving objects’ in videos will be fuzzy. So video de-noising can be transferred to problem of tracking and filtering along their trajectories. Three main steps exist in SMA, initial motion estimation, motion refinement and smooth, filtering along motion trajectory. Novel ‘Y’ detector is designed in the first step to accelerate searching procedure, where temporal and spatial vectors are considered. Besides, a jitter table is used to model singular motions. All possible vector candidates are minimized to a small but reliable set. In the second step, motion reliability is considered to deal with discontinuity of motion vector filed (MVF). Finally, a filter is adopted to eliminate irrelevant vectors in motion trajectory. The proposed filter can smooth MVs in an optimal way with maximum smoothness. Comparing with popular algorithms, SMA constraints searching region and prevent many complex transformations to a great extent. Moreover, it can be easily FPGA implemented. Numbers of experiments demonstrate that SMA is robust with good performance.

Scalable Motion Analysis Based Surveillance Video De-Noising

A low complexity but effective method in image and video detail enhancement – in-place similarity is proposed. In-place similarity is the statistic result of different kinds of nature image patches when similarity measurement is done in super resolution research, finding that patches in the upscale image have a good match around its original location in the corresponding lower ones. On the basis of the in-place similarity, a simple assumption is made that detail layer is the high-frequency component of twice search and patch match result. Then, images under the framework of super resolution techniques are enhanced. Unlike many current algorithms that need to adjust parameters for different images to acquire best outputs the approach mentioned here is adaptive. Moreover, many algorithms have outcome with intensity and contrast change, the algorithm proposed here can prevent images from over enhancement with obvious nature looking effect. Extensive experiments demonstrate that the algorithm mentioned here is simple with robust good performance both subjectively and objectively. It is particularly useful for practical applications and easily hardware implemented in FPGA, where input images require diverse detail textures.

In-place similarity and its applications in image and video detail enhancement

In this paper, we put forward a novel method in single image detail enhancement — residual homogeneity (RH). Residual homogeneity is a statistic result when we test different kinds of blocks in database, finding that blocks in residual images are with similar structure, that is, they are homogenous. Based on residual homogeneity, we make a simple assumption that detail layer of a single image is residual component after fast in-place search and block match. Then images are enhanced by applying framework of RH. Unlike many popular algorithms that need to adjust parameters by manual operation to get best performance, our approach is adaptive. Besides, many algorithms output images with intensity change, our RH can keep images from over enhancement with a nature look. RH is also robust to low bit rate H.265 system and runs faster than most popular methods. Moreover, it can be easily FPGA implemented as well. Numbers of experiments testify that our algorithm is robust with good performance both subjectively and objectively.

Single image detail enhancement via residual homogeneity

Handle everyday research tasks with reliable, citation-backed results

Your personal Research Agent to handle research tasks with citation-backed results

Popular Tasks used by Researchers

How can I help with your research?

Meet SciSpace

Get more enhanced response by uploading the PDFs you want me to reference.

No relevant PDFs in your library

SciSpace is the AI research assistant for academics. Run systematic literature reviews on 280M+ papers, and write papers with cited sources. Trusted by 1M+ students, PhDs & researchers.

SciSpace | AI for Research

Analyze PDFs

Code & Manuscripts

Funding & Grants

Literature & Patents

Medical & Clinical Data

Systematic Review

Visualize & Present

Web & Data

Build a Google Scholar-like website for your research.

Build a website

Create charts and images for your research

Create a Chart

Write a paper for submission to a journal

Draft a manuscript

Patent Search

Design eye-catching scientific posters in minutes.

Scientific Poster Generation

Systematic Literature Review

One task is running at the moment. Your messages will be shown right after.

Drag and drop or click here to browse

Loved by <highlight>1 million+</highlight> researchers

Extract a list of specific topics and their sources from unstructured text

Topics

Compare and analyze relevant papers that matches with your search

Papers

Get insights from PDFs and bookmarked papers from your library

My library

Recent searches

Try searching for:

Catch AI-generated content in scholarly and non-scholarly content

{ai} Detector

Ai Writer

Get PDF Summaries, highlighted text explanations 

Chat with PDF

Effortlessly create in-text citations and bibliographies in APA and 2,500 other formats

Citation generator

Get explanations, summaries, and answers on academic papers

Ease up your research workflow with {scispace}'s cohort of exciting AI tools

Elevate your academic writing skills and convey your ideas the way you want

Paraphraser

Explore our range of reading and writing tools

Your file is being prepared and should be ready in a few minutes. If it's a large file, it might take a bit longer. You can close this window, and we'll email you the file when it's done.

You have reached a maximum limit of <strong>{limit}</strong> columns in the table. Remove at least <strong>1</strong> column to add or create another one.

He Jiang

Author Tools

Chat about Author