A novel multi-scale violence and public gathering dataset for crowd behavior classification
<p dir="ltr">Dependable utilization of computer vision applications, such as smart surveillance, requires training deep learning networks on datasets that sufficiently represent the classes of interest. However, the bottleneck in many computer vision applications lies in the limited...
محفوظ في:
| المؤلف الرئيسي: | |
|---|---|
| مؤلفون آخرون: | , , |
| منشور في: |
2024
|
| الموضوعات: | |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
| _version_ | 1864513510714638336 |
|---|---|
| author | Almiqdad Elzein (13141038) |
| author2 | Emrah Basaran (19160743) Yin David Yang (19160746) Marwa Qaraqe (10135172) |
| author2_role | author author author |
| author_facet | Almiqdad Elzein (13141038) Emrah Basaran (19160743) Yin David Yang (19160746) Marwa Qaraqe (10135172) |
| author_role | author |
| dc.creator.none.fl_str_mv | Almiqdad Elzein (13141038) Emrah Basaran (19160743) Yin David Yang (19160746) Marwa Qaraqe (10135172) |
| dc.date.none.fl_str_mv | 2024-05-10T09:00:00Z |
| dc.identifier.none.fl_str_mv | 10.3389/fcomp.2024.1242690 |
| dc.relation.none.fl_str_mv | https://figshare.com/articles/journal_contribution/A_novel_multi-scale_violence_and_public_gathering_dataset_for_crowd_behavior_classification/26317000 |
| dc.rights.none.fl_str_mv | CC BY 4.0 info:eu-repo/semantics/openAccess |
| dc.subject.none.fl_str_mv | Information and computing sciences Computer vision and multimedia computation Data management and data science crowd analysis smart surveillance violence detection human action recognition computer vision |
| dc.title.none.fl_str_mv | A novel multi-scale violence and public gathering dataset for crowd behavior classification |
| dc.type.none.fl_str_mv | Text Journal contribution info:eu-repo/semantics/publishedVersion text contribution to journal |
| description | <p dir="ltr">Dependable utilization of computer vision applications, such as smart surveillance, requires training deep learning networks on datasets that sufficiently represent the classes of interest. However, the bottleneck in many computer vision applications lies in the limited availability of adequate datasets. One particular application that is of great importance for the safety of cities and crowded areas is smart surveillance. Conventional surveillance methods are reactive and often ineffective in enable real-time action. However, smart surveillance is a key component of smart and proactive security in a smart city. Motivated by a smart city application which aims at the automatic identification of concerning events for alerting law-enforcement and governmental agencies, we craft a large video dataset that focuses on the distinction between small-scale violence, large-scale violence, peaceful gatherings, and natural events. This dataset classifies public events along two axes, the size of the crowd observed and the level of perceived violence in the crowd. We name this newly-built dataset the Multi-Scale Violence and Public Gathering (<b>MSV-PG</b>) dataset. The videos in the dataset go through several pre-processing steps to prepare them to be fed into a deep learning architecture. We conduct several experiments on the <b>MSV-PG</b> datasets using a ResNet3D, a Swin Transformer and an R(2 + 1)D architecture. The results achieved by these models when trained on the <b>MSV-PG</b> dataset, 88.37%, 89.76%, and 89.3%, respectively, indicate that the dataset is well-labeled and is rich enough to train deep learning models for automatic smart surveillance for diverse scenarios.</p><h2>Other Information</h2><p dir="ltr">Published in: Frontiers in Computer Science<br>License: <a href="https://creativecommons.org/licenses/by/4.0/" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.3389/fcomp.2024.1242690" target="_blank">https://dx.doi.org/10.3389/fcomp.2024.1242690</a></p> |
| eu_rights_str_mv | openAccess |
| id | Manara2_cd711fc5b2fe5420f5e0dffb67bef0a9 |
| identifier_str_mv | 10.3389/fcomp.2024.1242690 |
| network_acronym_str | Manara2 |
| network_name_str | Manara2 |
| oai_identifier_str | oai:figshare.com:article/26317000 |
| publishDate | 2024 |
| repository.mail.fl_str_mv | |
| repository.name.fl_str_mv | |
| repository_id_str | |
| rights_invalid_str_mv | CC BY 4.0 |
| spelling | A novel multi-scale violence and public gathering dataset for crowd behavior classificationAlmiqdad Elzein (13141038)Emrah Basaran (19160743)Yin David Yang (19160746)Marwa Qaraqe (10135172)Information and computing sciencesComputer vision and multimedia computationData management and data sciencecrowd analysissmart surveillanceviolence detectionhuman action recognitioncomputer vision<p dir="ltr">Dependable utilization of computer vision applications, such as smart surveillance, requires training deep learning networks on datasets that sufficiently represent the classes of interest. However, the bottleneck in many computer vision applications lies in the limited availability of adequate datasets. One particular application that is of great importance for the safety of cities and crowded areas is smart surveillance. Conventional surveillance methods are reactive and often ineffective in enable real-time action. However, smart surveillance is a key component of smart and proactive security in a smart city. Motivated by a smart city application which aims at the automatic identification of concerning events for alerting law-enforcement and governmental agencies, we craft a large video dataset that focuses on the distinction between small-scale violence, large-scale violence, peaceful gatherings, and natural events. This dataset classifies public events along two axes, the size of the crowd observed and the level of perceived violence in the crowd. We name this newly-built dataset the Multi-Scale Violence and Public Gathering (<b>MSV-PG</b>) dataset. The videos in the dataset go through several pre-processing steps to prepare them to be fed into a deep learning architecture. We conduct several experiments on the <b>MSV-PG</b> datasets using a ResNet3D, a Swin Transformer and an R(2 + 1)D architecture. The results achieved by these models when trained on the <b>MSV-PG</b> dataset, 88.37%, 89.76%, and 89.3%, respectively, indicate that the dataset is well-labeled and is rich enough to train deep learning models for automatic smart surveillance for diverse scenarios.</p><h2>Other Information</h2><p dir="ltr">Published in: Frontiers in Computer Science<br>License: <a href="https://creativecommons.org/licenses/by/4.0/" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.3389/fcomp.2024.1242690" target="_blank">https://dx.doi.org/10.3389/fcomp.2024.1242690</a></p>2024-05-10T09:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.3389/fcomp.2024.1242690https://figshare.com/articles/journal_contribution/A_novel_multi-scale_violence_and_public_gathering_dataset_for_crowd_behavior_classification/26317000CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/263170002024-05-10T09:00:00Z |
| spellingShingle | A novel multi-scale violence and public gathering dataset for crowd behavior classification Almiqdad Elzein (13141038) Information and computing sciences Computer vision and multimedia computation Data management and data science crowd analysis smart surveillance violence detection human action recognition computer vision |
| status_str | publishedVersion |
| title | A novel multi-scale violence and public gathering dataset for crowd behavior classification |
| title_full | A novel multi-scale violence and public gathering dataset for crowd behavior classification |
| title_fullStr | A novel multi-scale violence and public gathering dataset for crowd behavior classification |
| title_full_unstemmed | A novel multi-scale violence and public gathering dataset for crowd behavior classification |
| title_short | A novel multi-scale violence and public gathering dataset for crowd behavior classification |
| title_sort | A novel multi-scale violence and public gathering dataset for crowd behavior classification |
| topic | Information and computing sciences Computer vision and multimedia computation Data management and data science crowd analysis smart surveillance violence detection human action recognition computer vision |