Learning an event-oriented and discriminative dictionary based on an adaptive label-consistent K-SVD method for event detection in soccer videos

Fakhar, Babak; Kanan, Hamidreza Rashidy; Behrad, Alireza

doi:10.1016/j.jvcir.2018.06.014

Cited by 10 publications

(5 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They annotated the soccer events by synchronizing the video clips and external text information (match reports) with coarse time constraints. Fakhar et al [24] presented a learning-based soccer event detection approach based on two main concepts. First, they analyzed the frame and estimated the saliency of each frame regarding soccer events.…”

Section: Low-level Features-based Event Recognition Approachesmentioning

confidence: 99%

AI-Driven Salient Soccer Events Recognition Framework for Next-Generation IoT-Enabled Environments

Muhammad

Ullah

Obaidat

et al. 2023

IEEE Internet Things J.

View full text Add to dashboard Cite

The salient events recognition of soccer matches in next-generation Internet of things (Nx-IoT) environment aims to analyze the performance of players/teams by the sports analytics and managerial staff. The embedded Nx-IoT devices carried by the soccer players during the match capture and transmit data to an Artificial Intelligence (AI)-assisted computing platform. The interconnectivity of data acquisition devices with an AI-assisted computing platform in the Nx-IoT environment will not only allow the spectators to track the formation of their favorite players during a soccer match but will also enable the managerial staff to evaluate the players' performance in the soccer match as well as in practice sessions. This Nx-IoT-enabled salient event detection feature can be provided to spectators and sports' managerial staff as a financial technology (FinTech) service. In this paper, we propose an efficient deep learning-based framework for multiperson salient soccer events recognition in IoT-enabled FinTech. The proposed framework performs event recognition in three steps: Firstly, image frames are extracted from video streams and resized in the preprocessing step to match the input of the deep network. Secondly, frame-level discriminative features are extracted using a pre-trained convolutional neural network (CNN) architecture. Thirdly, we employ a multi-layer long short-term memory (MLSTM) network to recognize high-level events in soccer videos by exploiting the sequential relation between adjacent frames. Moreover, we introduce a new soccer video events (SVE) dataset containing videos of six salient events of soccer game. To provide a strong baseline, we evaluate our newly created SVE dataset using different traditional machine learning and deep learning algorithms. We also perform event recognition on untrimmed soccer videos using our proposed framework and compare the results with state-of-the-art methods. The obtained results validate the suitability of our proposed framework for salient events recognition in Nx-IoT environments.

show abstract

Section: Low-level Features-based Event Recognition Approachesmentioning

confidence: 99%

AI-Driven Salient Soccer Events Recognition Framework for Next-Generation IoT-Enabled Environments

Muhammad

Ullah

Obaidat

et al. 2023

IEEE Internet Things J.

View full text Add to dashboard Cite

show abstract

“…Therefore, it is argued that the training data of normal and anomalous events can help an anomaly detection system learn better. To formulate a weakly-supervised learning approach, we resort to multiple instance learning [12,13]. Specifically, we propose to learn anomaly through a deep MIL framework by treating normal and anomalous surveillance videos as bags and short segments/clips of each video as instances in a bag.…”

Section: Literature Surveymentioning

confidence: 99%

Deep Surveillance System

2023

JSNDC

View full text Add to dashboard Cite

This study has been undertaken to investigate and implement multiple detection systems into a single surveillance system and check whether the input videos may comprise and capture a variety of realistic anomalies or not. In this paper, we propose to learn various anomalies by exploiting both normal and anomalous videos and implemented it to new model. Real time object detection is a vast, vibrant and sophisticated area of computer vision aimed towards object identification and recognition. Object detection detects the semantic objects of a class objects using Open source Computer Vision, which is a library of programming functions mainly trained towards real time computer vision in digital images and videos. The main aim behind this real time object detection is to help the peoples to overcome their difficulty. Real time object detection finds its uses in the areas like tracking objects, video surveillance, pedestrian detection, people counting, self-driving cars, face detection, tracking in sports and many more. This is achieved using Convolution, Probabilistic Neural Networks, etc. which are a representative tool of Deep learning. This project acts as an aiding tool for peoples who wants to take care of everything inside, outside, and around their house just for their full security expectations. Surveillance is a must for small houses to large-scale industries as they fulfil our safety aspects because theft and burglary have always been a problem. By combining this Surveillance idea to IoT and some Machine Learning stuff this will be a major product. The proposed project is a single autonomous surveillance system, based on analysis and detection technology. The proposed system is capable of monitoring all actions at once and alerts the concerned officials immediately and precisely.

show abstract

“…Source : (https://arxiv.org/pdf/1906.07538.pdf) artificial intelligence , the internet of things, and edge computing, and it shows considerable potential [4][5][6]. when used to video surveillance, EAI technology is an innovative and encouraging technique for transferring computational chores from the network's core to its periphery.…”

Section: Fig 1 Crowd Detectionmentioning

confidence: 99%

“…Several articles [9,15,16] have proposed a variety of distributed AI and Deep Learning (DL) methodologies for use with distributed computing clusters and cloud computing platforms. There are a lot of different ways that distributed AI algorithms and VS systems may be tested in different EC contexts [4,9,17]. During this investigation, we build a Distributed Intelligent Video Surveillance system that functions based on a distributed deep learning paradigm and then deploy it in an environment that consists of edge computing.…”

Section: Fig 1 Crowd Detectionmentioning

confidence: 99%

Big Data Enabled Realtime Crowd Surveillance and Threat Detection Using Artificial Intelligence and Deep Learning

Deepali Ujalambkar

2023

cienc.eng.

View full text Add to dashboard Cite

In recent years, security precautions at public events have become increasingly important as a direct response to the growth in the number of disruptive actions. There are several various varieties of closed-circuit televisions that are employed to perform 24-hour surveillance of public areas and the people that live such areas. Every person in a socialised society with a population of 1.6 billion is subjected to imprinted pictures an average of 30 times each day. Since the entire participation of the community is required to safeguard public areas from the most unexpected and lethal of events, it is difficult to discern whether an incident is an exceptional or casual occurrence because continual monitoring of human data makes it difficult to distinguish the difference. Within the scope of this study, we propose a method for identifying potentially threatening actions within footage obtained from closed-circuit television systems. To accomplish this objective, we must first extract individual still frames from the video and then examine the actions of the people seen in those still frames. We have placed a significant amount of reliance on both machine learning and deep learning algorithms to make this a reality. To automate this process, we must first develop a training model that makes use of many photos and a "Convolution Neural Network" that makes use of the Tensor Flow Python package. This model must be created before we can move on to automating the process. Every frame from every video that is provided will be used to train an algorithm that will analyse the film and evaluate whether it contains suspicious content or merely everyday activities. If we conclude that the activity was suspicious, the next phase of the study will concentrate on locating any weapons that may have been concealed on the corpse.

show abstract

Learning an event-oriented and discriminative dictionary based on an adaptive label-consistent K-SVD method for event detection in soccer videos

Cited by 10 publications

References 44 publications

AI-Driven Salient Soccer Events Recognition Framework for Next-Generation IoT-Enabled Environments

AI-Driven Salient Soccer Events Recognition Framework for Next-Generation IoT-Enabled Environments

Deep Surveillance System

Big Data Enabled Realtime Crowd Surveillance and Threat Detection Using Artificial Intelligence and Deep Learning

Contact Info

Product

Resources

About