A comprehensive list of data sources relating to violence against women and girls, bringing together a range of different sources from across government, academia and the voluntary sector.

This dataset has the following citation: M. Soliman, M. Kamal, M. Nashed, Y. Mostafa, B. Chawky, D. Khattab, “ Violence Recognition from Videos using Deep Learning Techniques”, Proc. 9th International Conference on Intelligent Computing and Information Systems (ICICIS'19), Cairo, pp. 79-84, 2019. please use it in case of using the dataset in research or engineering purpose ) when we start our Graduation Project Violence Recognition from Videos we found that there is shortage in available datasets related to violence between individuals so we decide to create new big dataset with variety of scenes

Content Our Dataset Contains 1000 Violence and 1000 non-violence videos collected from youtube videos, violence videos in our dataset contain many real street fights situations in several environments and conditions. also non-violence videos from our dataset are collected from many different human actions like sports, eating, walking …etc.

XD-Violence is a large-scale audio-visual dataset for violence detection in videos.

Dataset

The Bus Violence dataset is a large-scale collection of videos depicting violent and non-violent situations in public transport environments. This benchmark was gathered from multiple cameras located inside a moving bus where several people simulated violent actions, such as stealing an object from another person, fighting between passengers, etc. It contains 1,400 video clips manually annotated as having or not violent scenes, making it one of the biggest benchmarks for video violence detection in the literature.

Specifically, videos are recorded from three cameras at 25 Frames Per Second (FPS) --- two cameras located in the corners of the bus (with resolution 960x540 px) and one fisheye in the middle (1280x960 px). The clips have a minimum length of 16 frames and a maximum of 48 frames, capturing a very precise action (either violence or non-violence). The dataset is perfectly balanced, containing 700 videos of violence and 700 videos of non-violence.

The Bus Violence dataset is intended as a test data benchmark. However, for researchers interested in using our data also for training purposes, we provide training and test splits.

In this repository, we provide

• the 1,400 video clips divided into two folders named Violence /NoViolence, containing clips of violent situations and non-violent situations, respectively;

• two txt files containing the names of the videos belonging to the training and test splits, respectively.

Citing our work

If you found this dataset useful, please cite the following paper

@inproceedings{bus_violence_dataset_2022,
  title = {Bus Violence: An Open Benchmark for Video Violence Detection on Public Transport},
  doi = {10.3390/s22218345},
  url = {https://doi.org/10.3390%2Fs22218345},
  year = 2022,
  month = {oct},
  publisher = {{MDPI} {AG}},
  volume = {22},
  number = {21},
  pages = {8345},
  author = {Luca Ciampi and Pawe{\l} Foszner and Nicola Messina and Micha{\l} Staniszewski and Claudio Gennaro and Fabrizio Falchi and Gianluca Serao and Micha{\l} Cogiel and Dominik Golba and Agnieszka Szcz{\k{e}}sna and Giuseppe Amato},
  journal = {Sensors}
}

and this Zenodo Dataset

@dataset{pawel_bus_violence_zenodo,
  author    = {Paweł Foszner, Michał Staniszewski, Agnieszka Szczęsna, Michał Cogiel, Dominik Golba, Luca Ciampi, Nicola Messina, Claudio Gennaro, Fabrizio Falchi, Giuseppe Amato, Gianluca Serao},
  title    = {{Bus Violence: a large-scale benchmark for video violence detection in public transport}},
  month    = sep,
  year     = 2022,
  publisher  = {Zenodo},
  version   = {1.0.0},
  doi     = {10.5281/zenodo.7044203},
  url     = {https://doi.org/10.5281/zenodo.7044203}
}

Contact Information

Blees Sp. z o.o., Gliwice, Poland
mstaniszewski@blees.co

Acknowledgments

The presented dataset was supported by: European Union funds awarded to Blees Sp. z o.o. under grant POIR.01.01.01-00-0952/20-00 “Development of a system for analysing vision data captured by public transport vehicles interior monitoring, aimed at detecting undesirable situations/behaviours and passenger counting (including their classification by age group) and the objects they carry”); EC H2020 project "AI4media: a Centre of Excellence delivering next generation AI Research and Training at the service of Media, Society and Democracy" under GA 951911; research project INAROS (INtelligenza ARtificiale per il mOnitoraggio e Supporto agli anziani), Tuscany POR FSE CUP B53D21008060008.

License

The Bus Violence dataset was acquired by Blees Sp. z o.o. and is released under a Creative Commons Attribution license for non-commercial use.

Dataset for the paper "Not only Look, but also Listen: Learning Multimodal Violence Detection under Weak Supervision". The dataset is downloaded from the authors' website (https://roc-ng.github.io/XD-Violence/). Hosting this dataset on HuggingFace is just to make it easier for my own project to use this dataset. Please cite the original paper if you use this dataset.

This dataset contains aggregate data on violent index victimizations at the quarter level of each year (i.e., January – March, April – June, July – September, October – December), from 2001 to the present (1991 to present for Homicides), with a focus on those related to gun violence. Index crimes are 10 crime types selected by the FBI (codes 1-4) for special focus due to their seriousness and frequency. This dataset includes only those index crimes that involve bodily harm or the threat of bodily harm and are reported to the Chicago Police Department (CPD). Each row is aggregated up to victimization type, age group, sex, race, and whether the victimization was domestic-related. Aggregating at the quarter level provides large enough blocks of incidents to protect anonymity while allowing the end user to observe inter-year and intra-year variation. Any row where there were fewer than three incidents during a given quarter has been deleted to help prevent re-identification of victims. For example, if there were three domestic criminal sexual assaults during January to March 2020, all victims associated with those incidents have been removed from this dataset. Human trafficking victimizations have been aggregated separately due to the extremely small number of victimizations. This dataset includes a " GUNSHOT_INJURY_I " column to indicate whether the victimization involved a shooting, showing either Yes ("Y"), No ("N"), or Unknown ("UKNOWN.") For homicides, injury descriptions are available dating back to 1991, so the "shooting" column will read either "Y" or "N" to indicate whether the homicide was a fatal shooting or not. For non-fatal shootings, data is only available as of 2010. As a result, for any non-fatal shootings that occurred from 2010 to the present, the shooting column will read as “Y.” Non-fatal shooting victims will not be included in this dataset prior to 2010; they will be included in the authorized dataset, but with "UNKNOWN" in the shooting column. The dataset is refreshed daily, but excludes the most recent complete day to allow CPD time to gather the best available information. Each time the dataset is refreshed, records can change as CPD learns more about each victimization, especially those victimizations that are most recent. The data on the Mayor's Office Violence Reduction Dashboard is updated daily with an approximately 48-hour lag. As cases are passed from the initial reporting officer to the investigating detectives, some recorded data about incidents and victimizations may change once additional information arises. Regularly updated datasets on the City's public portal may change to reflect new or corrected information. How does this dataset classify victims? The methodology by which this dataset classifies victims of violent crime differs by victimization type: Homicide and non-fatal shooting victims: A victimization is considered a homicide victimization or non-fatal shooting victimization depending on its presence in CPD's homicide victims data table or its shooting victims data table. A victimization is considered a homicide only if it is present in CPD's homicide data table, while a victimization is considered a non-fatal shooting only if it is present in CPD's shooting data tables and absent from CPD's homicide data table. To determine the IUCR code of homicide and non-fatal shooting victimizations, we defer to the incident IUCR code available in CPD's Crimes, 2001-present dataset (available on the City's open data portal). If the IUCR code in CPD's Crimes dataset is inconsistent with the homicide/non-fatal shooting categorization, we defer to CPD's Victims dataset. For a criminal homicide, the only sensible IUCR codes are 0110 (first-degree murder) or 0130 (second-degree murder). For a non-fatal shooting, a sensible IUCR code must signify a criminal sexual assault, a robbery, or, most commonly, an aggravated battery. In rare instances, the IUCR code in CPD's Crimes and Vi

PhysicalViolence_dataset 2

## Overview

PhysicalViolence_dataset 2 is a dataset for object detection tasks - it contains Non Violence MQkX annotations for 3,937 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Annual data from the Crime Survey for England and Wales (CSEW), including when and where incidents happened and the victim's perception of the incident.

This dataset contains individual-level homicide and non-fatal shooting victimizations, including homicide data from 1991 to the present, and non-fatal shooting data from 2010 to the present (2010 is the earliest available year for shooting data). This dataset includes a "GUNSHOT_INJURY_I " column to indicate whether the victimization involved a shooting, showing either Yes ("Y"), No ("N"), or Unknown ("UKNOWN.") For homicides, injury descriptions are available dating back to 1991, so the "shooting" column will read either "Y" or "N" to indicate whether the homicide was a fatal shooting or not. For non-fatal shootings, data is only available as of 2010. As a result, for any non-fatal shootings that occurred from 2010 to the present, the shooting column will read as “Y.” Non-fatal shooting victims will not be included in this dataset prior to 2010; they will be included in the authorized-access dataset, but with "UNKNOWN" in the shooting column. Each row represents a single victimization, i.e., a unique event when an individual became the victim of a homicide or non-fatal shooting. Each row does not represent a unique victim—if someone is victimized multiple times there will be multiple rows for each of those distinct events. The dataset is refreshed daily, but excludes the most recent complete day to allow the Chicago Police Department (CPD) time to gather the best available information. Each time the dataset is refreshed, records can change as CPD learns more about each victimization, especially those victimizations that are most recent. The data on the Mayor's Office Violence Reduction Dashboard is updated daily with an approximately 48-hour lag. As cases are passed from the initial reporting officer to the investigating detectives, some recorded data about incidents and victimizations may change once additional information arises. Regularly updated datasets on the City's public portal may change to reflect new or corrected information. A version of this dataset with additional crime types is available by request. To make a request, please email dataportal@cityofchicago.org with the subject line: Violence Reduction Victims Access Request. Access will require an account on this site, which you may create at https://data.cityofchicago.org/signup. How does this dataset classify victims? The methodology by which this dataset classifies victims of violent crime differs by victimization type: Homicide and non-fatal shooting victims: A victimization is considered a homicide victimization or non-fatal shooting victimization depending on its presence in CPD's homicide victims data table or its shooting victims data table. A victimization is considered a homicide only if it is present in CPD's homicide data table, while a victimization is considered a non-fatal shooting only if it is present in CPD's shooting data tables and absent from CPD's homicide data table. To determine the IUCR code of homicide and non-fatal shooting victimizations, we defer to the incident IUCR code available in CPD's Crimes, 2001-present dataset (available on the City's open data portal). If the IUCR code in CPD's Crimes dataset is inconsistent with the homicide/non-fatal shooting categorization, we defer to CPD's Victims dataset. For a criminal homicide, the only sensible IUCR codes are 0110 (first-degree murder) or 0130 (second-degree murder). For a non-fatal shooting, a sensible IUCR code must signify a criminal sexual assault, a robbery, or, most commonly, an aggravated battery. In rare instances, the IUCR code in CPD's Crimes and Victims dataset do not align with the homicide/non-fatal shooting categorization: In instances where a homicide victimization does not correspond to an IUCR code 0110 or 0130, we set the IUCR code to "01XX" to indicate that the victimization was a homicide but we do not know whether it was a fi

The dataset was specifically created to address the need for violence detection in surveillance systems. It consists of self-recorded videos simulating different types of violent activities relevant to college environments. The dataset is organized into four distinct classes:SlapPunchKickGroup ViolenceOthers - Over Crowding, Loitering, Assault, AbuseEach video is labeled according to its corresponding class to facilitate supervised learning for violence detection models.

The National Intimate Partner and Sexual Violence Survey (NISVS) is an ongoing nationally representative survey that assessed experiences of sexual violence, stalking, and intimate partner violence among adult women and men in the United States and for each individual state. The survey focused exclusively on violence and collected information about: Sexual violence by any perpetrator, including information related to rape, being made to penetrate someone else, sexual coercion, unwanted sexual contact, and non-contact unwanted sexual experiences Stalking, including the use of newer technologies such as text messages, emails, monitoring devices (e.g., cameras and GPS, or global positioning system devices), by perpetrators known and unknown to the victim Physical violence by an intimate partner Psychological aggression by an intimate partner, including information on expressive forms of aggression and coercive control Control of reproductive or sexual health by an intimate partner The NISVS project and data collection was overseen by the Centers for Disease and Prevention (CDC). The overall cost of the NISVS project was shared between the CDC, the Department of Defense (DoD) and the National Institute of Justice (NIJ). ICPSR has multiple versions of NISVS data that you can access by clicking on the links provided below.

Due to structural racism and social policies that perpetuate inequality, violence disproportionately affects some communities and leads to outcomes that results in higher rates of school bullying, homicides, and firearm death rates.

Domestic violence data from the Washington Association of Sheriffs and Police Chiefs (WASPC).

This table regroups a series of indicators related to violence against women collected from various sources (national surveys, international databases).

Find more Pacific data on PDH.stat.

Detected Images Violence

## Overview

Detected Images Violence is a dataset for classification tasks - it contains Violent Images annotations for 9,909 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

This study contains data on the attitudes of 1,374 American men aged 16-64 toward violence in 1969. The study was undertaken to examine the levels of violence that can be viewed as justified to bring about social control or social change. Also emphasized were the role of the respondents' personal values, their definitions of violence, and their identification with the groups involved in violence. Some of the open-ended questions in the structured interview probed the respondents' general concerns, their attitudes toward violence, and their views on the causes of and ways of preventing violence. In questions grouped into categories of "violence for social control" and "violence for social change", respondents were asked to react to situations involving protests and other disturbances such as hoodlum gang disturbances, students' protests, and Black protest demonstrations. Repondents' opinions were sought on the appropriate police actions in these situations and the frequency with which certain control measures should be utilized. Respondents were also asked in three different situations whether they believed change could be effected without action involving property damage or injury, or if change could only be effected with protests in which some people were killed. Demographic variables describe age, sex, date of birth, nationality, occupation, education, religion, and family income. A supplementary sample of Black men is also included in this study in order to permit separate analysis on the basis of race.

The Gun Violence Corpus (GVC) consists of 241 unique incidents for which we have structured data on a) location, b) time c) the name, gender and age of the victims and d) the status of the victims after the incident: killed or injured. For these data, 510 news articles were gathered following the 'data to text' approach. The structured data and articles report on a variety of gun violence incidents, such as drive-by shootings, murder-suicides, hunting accidents, involuntary gun discharges, etcetera. The documents have been manually annotated for all mentions that make reference to the gun violence incident at hand.

These data are part of NACJD's Fast Track Release and are distributed as they were received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except for the removal of direct identifiers. Users should refer to the accompanying readme file for a brief description of the files available with this collection and consult the investigator(s) if further information is needed. The study investigated the feasibility of implementing the Cardiff Model. The Cardiff Model is a unique violence surveillance system and intervention that involves data sharing and violence prevention planning between law enforcement and the medical field. Anonymized data on assaults from emergency and police departments (EDs; PDs) are combined to detail assault incidents and "hotspots." Data are discussed by a multidisciplinary consortium, which develops and implements a data-informed violence prevention action plan that includes behavioral, environmental, and policy changes to impact violence. Model actions led to decreases in injurious assaults and this model is now statutory in the United Kingdom. The Cardiff Model has never been translated to the U.S. and would require an investigation within our health care system and in different geographical and population contexts. This study investigated the feasibility of essential Cardiff Model Components in order to refine study procedures and situate this community to request further funds for full model implementation. As part of this study, researchers collected a number of feasibility measures from ED and study staff to evaluate the feasibility of translating included model components. Geospatial and statistical analyses investigated the added benefit of the combined ED, PD and Emergency Medical Services (EMS) data. The study contains 1 SPSS data files (CHW Data_1.1.15 to 7.31.16.sav (n=748; 14 variables)), 1 STATA data file (nurse survey data.dta (n=43; 26 variables)), a text document (Nurse Survey_Qualitative data.txt), and 1 excel file (CHW Incidents_Block level data only.xlsx).

This dataset provides a breakdown of the City of Boston's efforts to fund Community-based Organizations that do violence prevention, intervention, and recovery work within the City. It highlights the different cabinets/bureaus/departments that offer funding related to community violence prevention, the funded organizations and programs, the amount of funding that the city has offered, and the ways to get in contact with each of the funded organizations.