PRESS-NEWS.org - Press Release Distribution
PRESS RELEASES DISTRIBUTION

COVAD: Content-oriented video anomaly detection using a self attention-based deep learning model

COVAD: Content-oriented video anomaly detection using a self attention-based deep learning model
2023-05-09
(Press-News.org)

Video anomaly detection is a research hotspot in the field of computer vision, attracting many researchers.Video anomaly detection differs from traditional video analysis. Usually, abnormal events occur only in a small percentage of the video pixels and therefore, it is unnecessary to focus on all the video pixels as most of

them are harmless—called “the background”. Therefore, in the video feature extraction process, attention should be focused on a few detectable partial objects. Object detection is very complicated and consumes a significant amount of time during video processing. Therefore, it is not advisable to use object detection in the

training phase to focus attention on the anomalous parts.

In this paper, a content-based video anomaly detection algorithm (COVAD) is proposed, and its network structure is modified based on the original memory-based video anomaly detection algorithm. The main goal of optimization in the training network is to focus on the objects in the video frame. We used a content-based attention mechanism to optimize the structure of the encoding network and remove the last batch of the normalization layer of the U-Net network. The former is used to focus on the target or content in the video and the latter is used to limit the powerful bias of the neural network because it is important to blur the boundary between normal and abnormal data in powerful representations. Compared with the object detection algorithm, the attention mechanism is lightweight, does not consume a lot of time, and can effectively process videos. The memory storage module stores more important content information than the entire video frame pixels. Our experiments were deployed on the USCD and Avenue datasets, and the experimental results show that the proposed algorithm has better results than the benchmark models.

The main contributions of this paper are: 1) to propose a novel video anomaly detection method—called COVAD—for future frame prediction by combining the content-based attention mechanism, which can resist the interference of noise and focus on extracting the features of objects in the video; 2) to redefine the memory module that is used to classify and memorize various normal behavioral patterns available in video streams; and 3) to further improve the performance of video anomaly detection models focused on both normal and exceptional events. The experimental results show that the performance of the proposed COVAD algorithm is significantly higher than that of the baseline models considered in this paper.

END


[Attachments] See images for this press release:
COVAD: Content-oriented video anomaly detection using a self attention-based deep learning model COVAD: Content-oriented video anomaly detection using a self attention-based deep learning model 2 COVAD: Content-oriented video anomaly detection using a self attention-based deep learning model 3

ELSE PRESS RELEASES FROM THIS DATE:

New technique enables in-vivo analysis of protein complexes

2023-05-09
As the executor of life activities, proteins exert their specific biological functions through interactions such as forming protein complexes. The localization effects, crowding effects, and organelle microenvironments within cells are crucial for maintaining the structure and function of protein complexes. Recently, a research team led by Prof. ZHANG Lihua from the Dalian Institute of Chemical Physics (DICP) of the Chinese Academy of Sciences (CAS) has developed a glycosidic-bond-based mass-spectrometry-cleavable cross-linker, which improves the data ...

Scientists raise concerns about popular COVID disinfectants

2023-05-09
The COVID-19 pandemic has boosted the unnecessary use of antimicrobial chemicals linked to health problems, antimicrobial resistance, and environmental harm, warn more than two dozen scientists in the peer-reviewed journal Environmental Science & Technology. Their critical review details how quaternary ammonium compounds (QACs) are increasingly marketed and used in home, healthcare, education, and workplace settings despite the availability of safer alternatives and in some cases limited evidence of reduced disease transmission. “Disinfectant wipes containing ...

Virginia Tech, George Mason University partner to develop networking infrastructure for satellite constellations

Virginia Tech, George Mason University partner to develop networking infrastructure for satellite constellations
2023-05-09
The race is on to provide high speed satellite internet to the Earth’s most remote areas. New tech companies such as Starlink, One Web, and Amazon’s Kuiper are competing with traditional, established “satcomm” companies such as Thuraya and Inmarsat to provide global high speed, low latency satellite internet across the globe. These new mega-constellations rely on tens of thousands of small low earth orbit satellites flying at a few hundred miles altitude. Network interconnectivity is a basic building block for providing the fastest, more reliable coverage to end users. While all these mega-constellations are driven to provide high ...

Lifu Huang receives NSF CAREER award to lay new ground for information extraction without relying on humans

Lifu Huang receives NSF CAREER award to lay new ground for information extraction without relying on humans
2023-05-08
Considering the millions of research papers and reports from open domains such as biomedicine, agriculture, and manufacturing, it is humanly impossible to keep up with all the findings. Constantly emerging world events present a similar challenge because they are difficult to track and even harder to analyze without looking into thousands of articles.  To address the problem of relying on human effort in situations such as these, Lifu Huang, an assistant professor in the Department of Computer Science and core faculty at the Sanghani Center for Artificial Intelligence ...

LY6 gene family: potential tumor antigens and prognostic biomarkers in endometrial cancer

LY6 gene family: potential tumor antigens and prognostic biomarkers in endometrial cancer
2023-05-08
“Importantly, the expression of several LY6 genes is elevated in UCEC [uterine corpus endometrial carcinoma] when compared to the expression in normal uterine tissue.” BUFFALO, NY- May 8, 2023 – A new research paper was published in Oncotarget's Volume 14 on May 4, 2023, entitled, “Human LY6 gene family: potential tumor-associated antigens and biomarkers of prognosis in uterine corpus endometrial carcinoma.” The human Lymphocyte antigen-6 (LY6) gene family has recently gained interest for its possible role in tumor progression. In this new study, researchers Luke A. Rathbun, Anthony M. Magliocco ...

Atmospheric research provides clear evidence of human-caused climate change signal associated with CO2 increases

2023-05-08
Woods Hole, Mass. (Monday, May 8, 2023) -- New research provides clear evidence of a human “fingerprint” on climate change and shows that specific signals from human activities have altered the temperature structure of Earth’s atmosphere. Differences between tropospheric and lower stratospheric temperature trends have long been recognized as a fingerprint of human effects on climate. This fingerprint, however, neglected information from the mid to upper stratosphere, 25 to 50 kilometers above the Earth’s surface. “Including this information improves the detectability of a human fingerprint by a factor of five. Enhanced detectability occurs because the mid ...

UC Irvine, NASA JPL researchers discover a cause of rapid ice melting in Greenland

2023-05-08
Irvine, Calif., May 8, 2023 – While conducting a study of Petermann Glacier in northwest Greenland, researchers at the University of California, Irvine and NASA’s Jet Propulsion Laboratory uncovered a previously unseen way in which the ice and ocean interact. The glaciologists said their findings could mean that the climate community has been vastly underestimating the magnitude of future sea level rise caused by polar ice deterioration. Using satellite radar data from three European missions, the UCI/NASA team learned that Petermann Glacier’s grounding line – where ice detaches from the land bed and begins floating in the ocean – shifts substantially ...

A new twist on chirality: researchers extend the concept of directionality and propose a new class of materials

A new twist on chirality: researchers extend the concept of directionality and propose a new class of materials
2023-05-08
It is often desirable to restrict flows—whether of sound, electricity, or heat—to one direction, but naturally occurring systems almost never allow this. However, unidirectional flow can indeed be engineered under certain conditions, and the resulting systems are said to exhibit chiral behavior. The concept of chirality is traditionally limited to single direction flows in one dimension. In 2021, however, researchers working with Taylor Hughes, a professor of physics at the University of Illinois Urbana-Champaign, ...

New York’s fertility rate drops, average age of mothers rises

2023-05-08
ITHACA, NY – A decline in New York’s childbirth rate is showing no sign of reversing and many women are waiting longer to have children, according to newly compiled data from the Program in Applied Demographics (PAD) in the Cornell Jeb E. Brooks School of Public Policy. In 2011, about 241,312 were born in New York. In 2021, that number was 210,742 – a 13% decline. New York state’s total fertility rate (TFR) – the average number of births a woman would have in her lifetime if current patterns continue – dropped from ...

Department of Energy announces $45 million for Inertial Fusion Energy (IFE)

2023-05-08
WASHINGTON, D.C. – Today, at the celebration ceremony of the historic achievement of fusion ignition at the National Ignition Facility (NIF), the U.S. Secretary of Energy Jennifer Granholm announced a plan to provide up to $45 million to support Inertial Fusion Energy (IFE) research and development. Fusion, the process that powers the sun and stars, has the potential to provide clean, safe, and reliable carbon-free energy on earth. Harnessing fusion energy is one of the greatest scientific and technological challenges of the 21st century. Fusion requires the fuel to ...

LAST 30 PRESS RELEASES:

Genetic markers for depression reveal consistent patterns in psychiatric treatment outcomes

Deep-sea mining could harm remote ocean ecosystems

Stanford researcher develops machine learning models to decode brain aging at cellular level

AI shortens the development time of new materials

Insulin on edge: Study identifies stress-triggered gene behind diabetes

Wildlife forensic scientists develop new tool to detect elephant ivory disguised as legal mammoth ivory

Organ preservation strategies: Extended sleeve lobectomy after neoadjuvant immunochemotherapy offers optimal option for centrally located NSCLC

Doubts cast over suggestions incestuous ‘god-kings’ ruled during Neolithic Ireland

Interpretation on feature groups for tree models

Military discharge is a time of challenge and opportunity

Common pregnancy complications may be a signal of future stroke risk

Barcodes uncover early blueprints of our cellular origins

Stanford Medicine-led phase 3 trial shows gene therapy skin grafts help epidermolysis bullosa

‘Pill-on-a-thread’ could replace endoscopies for half of all patients being monitored for esophageal cancer risk

Study casts doubt on ‘incestuous royalty’ in Neolithic Ireland

Heart valve developed at UC Irvine shines in early-stage preclinical testing

In diseases due to exposure to toxic particles like gout, macrophages elicit separate pathways for inflammation and lysosomal function

Zoning out could be beneficial—and may actually help us learn faster

Weekly semaglutide improves blood sugar and weight in adults with Type 1 diabetes

Concerned father, statistician develops software to improve skills therapy

Your smartwatch might know you’re sick before you do — and it might help stop pandemics

ImmunoPET tracer enhances early detection of liver cancer

AI-based brain-mapping software receives FDA market authorization

New PET tracer identifies diverse invasive mold infections behind life-threatening illnesses in cancer and transplant patients

Current Pharmaceutical Analysis (CPA) achieves notable impact factor growth in latest journal citation reports

AI chatbot safeguards fail to prevent spread of health disinformation

UTIA researcher to receive award from the Soil and Water Conservation Society

HSE linguists study how bilinguals use phrases with numerals in Russian

Cold winters halt the northward spread of species in a warming climate

Study finds early signs of widespread coastal marsh decline

[Press-News.org] COVAD: Content-oriented video anomaly detection using a self attention-based deep learning model