Crowd anomaly detection using Aggregation of Ensembles of fine-tuned ConvNets

Singh, Kulbir; Rajora, Shantanu; Vishwakarma, Dinesh Kumar; Tripathi, Gaurav; Kumar, Sandeep; Walia, Gurjit Singh

doi:10.1016/j.neucom.2019.08.059

Cited by 94 publications

(40 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In [8] , an ensemble of several models was proposed. Three different models of CNN were fine tuned as feature extractors (namely VGGNet, AlexNet and GoogLeNet), and their outputs were concatenated to form a feature vector.…”

Section: Deep Learning For Crowd Anomaly Detection: Approaches and Numentioning

confidence: 99%

See 1 more Smart Citation

Revisiting crowd behaviour analysis through deep learning: Taxonomy, anomaly detection, crowd emotions, datasets, opportunities and prospects

Sánchez

Hupont²,

Tabik

et al. 2020

Information Fusion

View full text Add to dashboard Cite

Crowd behaviour analysis is an emerging research area. Due to its novelty, a proper taxonomy to organise its different sub-tasks is still missing. This paper proposes a taxonomic organisation of existing works following a pipeline, where sub-problems in last stages benefit from the results in previous ones. Models that employ Deep Learning to solve crowd anomaly detection, one of the proposed stages, are reviewed in depth, and the few works that address emotional aspects of crowds are outlined. The importance of bringing emotional aspects into the study of crowd behaviour is remarked, together with the necessity of producing real-world, challenging datasets in order to improve the current solutions. Opportunities for fusing these models into already functioning video analytics systems are proposed.

show abstract

Section: Deep Learning For Crowd Anomaly Detection: Approaches and Numentioning

confidence: 99%

“…A large number of publications have addressed crowd behaviour analysis using Deep Learning techniques in their pipelines [6] , [7] , [8] . Nevertheless, most of these works are sparse and difficult to compare.…”

Section: Introductionmentioning

confidence: 99%

Revisiting crowd behaviour analysis through deep learning: Taxonomy, anomaly detection, crowd emotions, datasets, opportunities and prospects

Sánchez

Hupont²,

Tabik

et al. 2020

Information Fusion

View full text Add to dashboard Cite

show abstract

“…Ensemble methods have been utilized to make robust predictions, especially in the semisupervised anomaly detection to boost base classifier output; bagging and boosting are mostly used with traditional methodologies or neural networks such as autoencoders [9], [10]. A recent approach utilizes an ensemble of multiple finetuned CNNs to learn different abstractions of features from crowd anomaly videos, known as Aggregation of Ensembles (AOE), to achieve learning of nuanced differences between normal and abnormal events [39]. However, using the same dataset samples over multiple CNNs is prone to feature redundancy.…”

Section: Related Workmentioning

confidence: 99%

IBaggedFCNet: An Ensemble Framework for Anomaly Detection in Surveillance Videos

et al. 2020

View full text Add to dashboard Cite

“…Um AOE (Aggregation of Ensembles), utilizando redes convolucionais CNN (Convolutional Neural Network ) prétreinadas e um conjunto de classificadores, foi proposto por Singh et al (2020). Sua abordagem foi inspirada no conceito de que um conjunto de diferentes CNN ajustadas, representa vários níveis de semântica e, portanto, codificam um amplo conjunto de características robustas.…”

Section: Trabalhos Relacionadosunclassified

“…Naárea de segurança pública, os sistemas de visão computacional têm sido utilizados para detectar atividades anômalas, ou mesmo suspeitas (Ravanbakhsh et al, 2019;Singh et al, 2020). Anomalias são padrões em dados que Trabalho realizado com apoio da FAPES -Fundação de Amparo a Pesquisa e Inovação do Espırito Santo através do Projeto 577/2018, e da NVIDIA Corporation através da doação da GPU Titan V.…”

Section: Introductionunclassified

Detecção de Anomalias em Vias Públicas Usando Características Espaciais e um Classificador Sequencial Bidirecional

Bento

Vassallo

Samatelo

2020

Anais Do Congresso Brasileiro De Automática 2020

View full text Add to dashboard Cite

Detecção de anomalias consiste na identificação de eventos que não estão em conformidade com um padrão de comportamento esperado. No contexto de segurança em vias públicas, a detecção automática de eventos anômalos através de video, tem aplicação na identificação de comportamentos suspeitos. Nesse artigo é proposta uma abordagem para o problema da detecção automática de eventos anômalos em vı́deos de vias públicas baseado em um modelo de redes neurais profundas end-to-end, composto de duas partes: um extrator de caracterı́sticas espaciais baseado em uma rede neural convolucional pre-treinada, e um classificador de sequências temporais baseado em camadas recorrentes empilhadas. Realizamos experimentos nos conjuntos de dados UCSD Anomaly Detection Dataset. Os resultados foram avaliados com as métricas Area Under the Receiver Operating Characteristic Curve - AUC, Area Under the Precision vs Recall Curve - AUPRC e Equal Error Rate – EER. Durante os experimentos, o modelo obteve AUC acima de 95% e EER abaixo de 9%, os quais são resultados compatı́veis com a literatura atual.

show abstract

Crowd anomaly detection using Aggregation of Ensembles of fine-tuned ConvNets

Cited by 94 publications

References 21 publications

Revisiting crowd behaviour analysis through deep learning: Taxonomy, anomaly detection, crowd emotions, datasets, opportunities and prospects

Revisiting crowd behaviour analysis through deep learning: Taxonomy, anomaly detection, crowd emotions, datasets, opportunities and prospects

IBaggedFCNet: An Ensemble Framework for Anomaly Detection in Surveillance Videos

Detecção de Anomalias em Vias Públicas Usando Características Espaciais e um Classificador Sequencial Bidirecional

Contact Info

Product

Resources

About