This study was undertaken to obtain information on the characteristics of gun ownership, gun-carrying practices, and weapons-related incidents in the United States -- specifically, gun use and other weapons used in self-defense against humans and animals. Data were gathered using a national random-digit-dial telephone survey. The respondents were comprised of 1,905 randomly-selected adults aged 18 and older living in the 50 United States. All interviews were completed between May 28 and July 2, 1996. The sample was designed to be a representative sample of households, not of individuals, so researchers did not interview more than one adult from each household. To start the interview, six qualifying questions were asked, dealing with (1) gun ownership, (2) gun-carrying practices, (3) gun display against the respondent, (4) gun use in self-defense against animals, (5) gun use in self-defense against people, and (6) other weapons used in self-defense. A "yes" response to a qualifying question led to a series of additional questions on the same topic as the qualifying question. Part 1, Survey Data, contains the coded data obtained during the interviews, and Part 2, Open-Ended-Verbatim Responses, consists of the answers to open-ended questions provided by the respondents. Information collected for Part 1 covers how many firearms were owned by household members, types of firearms owned (handguns, revolvers, pistols, fully automatic weapons, and assault weapons), whether the respondent personally owned a gun, reasons for owning a gun, type of gun carried, whether the gun was ever kept loaded, kept concealed, used for personal protection, or used for work, and whether the respondent had a permit to carry the gun. Additional questions focused on incidents in which a gun was displayed in a hostile manner against the respondent, including the number of times such an incident took place, the _location of the event in which the gun was displayed against the respondent, whether the police were contacted, whether the individual displaying the gun was known to the respondent, whether the incident was a burglary, robbery, or other planned assault, and the number of shots fired during the incident. Variables concerning gun use by the respondent in self-defense against an animal include the number of times the respondent used a gun in this manner and whether the respondent was hunting at the time of the incident. Other variables in Part 1 deal with gun use in self-defense against people, such as the _location of the event, if the other individual knew the respondent had a gun, the type of gun used, any injuries to the respondent or to the individual that required medical attention or hospitalization, whether the incident was reported to the police, whether there were any arrests, whether other weapons were used in self-defense, the type of other weapon used, _location of the incident in which the other weapon was used, and whether the respondent was working as a police officer or security guard or was in the military at the time of the event. Demographic variables in Part 1 include the gender, race, age, household income, and type of community (city, suburb, or rural) in which the respondent lived. Open-ended questions asked during the interview comprise the variables in Part 2. Responses include descriptions of where the respondent was when he or she displayed a gun (in self-defense or otherwise), specific reasons why the respondent displayed a gun, how the other individual reacted when the respondent displayed the gun, how the individual knew the respondent had a gun, whether the police were contacted for specific self-defense events, and if not, why not.

The share of American households owning at least one firearm has remained relatively steady since 1972, hovering between ** percent and ** percent. In 2024, about ** percent of U.S. households had at least one gun in their possession. Additional information on firearms in the United States Firearms command a higher degree of cultural significance in the United States than any other country in the world. Since the inclusion of the right to bear arms in the Second Amendment to the Constitution of the United States, firearms have held symbolic power beyond their already obvious material power. Despite many Americans being proud gun-owners, a large movement exists within the country in opposition to the freedom afforded to those in possession of these potentially deadly weapons. Those opposed to current gun regulation have sourced their anger from the large number of deaths due to firearms in the country, as well as the high frequency of gun violence apparent in comparison to other developed countries. Furthermore, the United States has fallen victim to a number of mass shootings in the last two decades, most of which have raised questions over the ease at which a person can obtain a firearm. Although this movement holds a significant position in the public political discourse of the United States, meaningful change regarding the legislation dictating the ownership of firearms has not occurred. Critics have pointed to the influence possessed by the National Rifle Association through their lobbying of public officials. The National Rifle Association also lobbies for the interests of firearm manufacturing in the United States, which has continued to rise since a fall in the early 2000s.

Here are a few use cases for this project:

1. Security Surveillance: This model can be used in CCTV and other video monitoring systems for real-time identification of people carrying weapons. It could prove vital for public safety in locations such as airports, stations, or public squares by detecting potential threats and avoiding incidents.

2. Smart Policing: Law enforcement can use this model in police body cameras or drones to help detect individuals with weapons during crowded events, protests, or routine patrols, thereby aiding in maintaining law and order.

3. Video Analytics: The model can be utilized in analysing recorded videos for forensics and investigation purposes. For example, law enforcement could use the model to quickly analyze surveillance footage from a crime scene to identify suspects carrying weapons.

4. Gaming and Entertainment: In the gaming and movie industry, the algorithm can be helpful in creating effects, animation, or real-time simulation where the interaction of the person with a weapon is required.

5. Virtual Training Simulators: This model could be used in military or police training simulators, helping to create a real-world, responsive environment where trainees interact with virtual characters carrying weapons.

Texas was the state with the highest number of registered weapons in the United States in 2024, with 1,136,732 firearms. Rhode Island, on the other hand, had the least, with 4,895 registered firearms. Gun laws in the United States Gun ownership in the U.S. is protected by the 2nd Amendment of the Constitution, which allows citizens to own firearms and form a militia if necessary. Outside of the 2nd Amendment, gun laws in the U.S. vary from state to state, and gun owners are subject to the laws of the state they are currently in, not necessarily the state they live in. For example, if concealed carry is allowed in a gun owner’s state of residence but not in the state they are traveling in, the owner is subject to the law of the state they are traveling in. Civilian-owned firearms The United States is estimated to have the highest rate of civilian-owned firearms in the world, more than double that of Yemen, which has the second-highest gun ownership rate. Unfortunately, along with high gun ownership rates comes a higher number of homicides by firearm, which was about 13,529 homicides in 2023.

People And Weapons

## Overview

People And Weapons is a dataset for object detection tasks - it contains People annotations for 4,365 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 project investigated the experiences of New York City youth ages 16-24 who were at high risk for gun violence (e.g., carried a gun, been shot or shot at). Youth participants were recruited from three neighborhoods with historically high rates of gun violence when compared to the city as a whole--Brownsville (Brooklyn), Morrisania (Bronx), and East Harlem (Manhattan). This study explores the complex confluence of individual, situational, and environmental factors that influence youth gun acquisition and use. This study is part of a broader effort to build an evidence-based foundation for individual and community interventions, and policies that will more effectively support these young people and prevent youth gun violence. Through interviews with 330 youth, this study seeks to answer these questions: What are the reasons young people carry guns? How do young people talk about having and using guns? What are young people's social networks like, and what roles do guns play in thesenetworks? Interviews covered the following topics: neighborhood perceptions; perceptions of and experiences with the police, gangs, guns, and violence; substance use; criminal history; and demographics: race, gender, age, legal status, relationship status, living situation, _location, number of children, drug use, and education.

Here are a few use cases for this project:

1. Security Surveillance: This model can be used in real-time security surveillance systems in various high-risk areas such as banks, airports, schools, or government buildings to detect any individuals possessing firearms. It can alert security personnel immediately once a firearm is detected.

2. Crime Prevention: Law enforcement agencies and police departments can use this model to identify and track individuals carrying guns in public spaces, helping to prevent potential violent incidents and crimes.

3. Social Media Monitoring: The model can be applied to scan and monitor social media content and online videos for illegal display of firearms, significantly aiding in online safety and cyber law enforcement.

4. Video Content Filtering: Online platforms such as YouTube or TikTok can leverage this model to filter out or flag content that contains guns, ensuring that such content doesn't reach the wrong audience like minors or individuals susceptible to violence.

5. AI-based Video Games: Developers may use this model in AI-enhanced video games for accurately identifying guns and players, contributing to more realistic game scenarios or possibly supporting anti-cheating mechanisms.

Information from Bloomington Police Department regarding guns reported stolen. Key code for Race: A- Asian/Pacific Island, Non-Hispanic B- African American, Non-Hispanic C- Hawaiian/Other Pacific Island, Hispanic H- Hawaiian/Other Pacific Island, Non-Hispanic I- Indian/Alaskan Native, Non-Hispanic K- African American, Hispanic L- Caucasian, Hispanic N- Indian/Alaskan Native, Hispanic P- Asian/Pacific Island, Hispanic S- Asian, Non-Hispanic T- Asian, Hispanic U- Unknown W- Caucasian, Non-Hispanic Key Code for Reading Districts: Example: LB519 L for Law call or incident B stands for Bloomington 5 is the district or beat where incident occurred All numbers following represents a grid sector. Disclaimer: The Bloomington Police Department takes great effort in making open data as accurate as possible, but there is no avoiding the introduction of errors in this process, which relies on data provided by many people and that cannot always be verified. Information contained in this dataset may change over a period of time. The Bloomington Police Department is not responsible for any error or omission from this data, or for the use or interpretation of the results of any research conducted.

Number and percentage of homicide victims, by type of firearm used to commit the homicide (total firearms; handgun; rifle or shotgun; fully automatic firearm; sawed-off rifle or shotgun; firearm-like weapons; other firearms, type unknown), Canada, 1974 to 2018.

Data tables relating to offences involving weapons as recorded by police and hospital episode statistics.

Gun ownership is a highly a consequential political behavior. It often signifies a belief about the inadequacy of state-provided security and leads to membership in a powerful political constituency. As a result, it is important to understand why people buy guns and how shifting purchasing patterns affect the composition of the broader gun owning community. We address these topics by exploring the dynamics of the gun-buying spike that took place during the COVID-19 pandemic, which was one of the largest in American history. We find that feelings of diffuse threat prompted many individuals to buy guns. Moreover, we show that new gun owners, even more than buyers who already owned guns, exhibit strong conspiracy and anti-system beliefs. These findings have substantial consequences for the subsequent population of gun owners and provide insight into how social disruptions can alter the nature of political groups.

People With Gun (Thai)

## Overview

People With Gun (Thai) is a dataset for object detection tasks - it contains Pistol annotations for 244 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).

The Concealed Carry Weapons License Database (CCWLD) is a longitudinal collection of state and county-level data on concealed carry weapons licenses (CCWs). Data were collected from a series of internet searches and freedom of information requests sent to state governments during the fall of 2019 and winter of 2020. Data cleaning was conducted by research assistants and by Trent Steidley in the winter of 2021.Documentation memos for each state are provided in the archived files. Along with raw data files, Stata syntax for cleaning, and the final cleaned database.This database was supported with funding from the Center on American Politics at the University of Denver and a Professional Research Opportunity for Faculty (PROF) grant from the University of Denver. If you use these data in your research please cite them appropriately.

This dataset consists of gun violence within Jefferson county that may fall within LMPDs radar, including non-fatal shootings, homicides, as well as shot-spotter data. The mapping data points where there are victims have been obfuscated to maintain privacy, while still being accurate enough to be placed in its correct boundaries, particularly around, neighborhoods, ZIP Codes, Council districts, and police divisions. The data also excludes any victim information that could be used to identify any individual. this data is used to make the public aware of what is going on in their communities. The data consists of only criminal incidents, excluding any cases that are deemed non-criminal.Field NameField DescriptionCase numberPolice report number. For ShotSpotter detections, it is the ShotSpotter ID.DateTimeDate and time in which the original incident occurred. Time is rounded down.AddressAddress rounded down to the one hundred block of where the initial incident occured. Unless it is an intersection.NeighborhoodNeighborhood in which the original incident occurred.Council DistrictCouncil district in which the original incident occurred.LatitudeLatitude coordinate used to map the incidentLongitudeLongitude coordinate used to map the incidentZIP CodeZIP Code in which the original incident occurred.Crime Typea distinction between incidents, whether it is a non-fatal shooting, homicide, or a ShotSpotter detection.CauseUsed to differentiate on the cause of death for homicide victims.SexGender of the victim of the initial incident.RaceRace/Ethnicity of the victim in a given incident.Age GroupCategorized age groups used to anonymize victim information.Division NamePolice division or department where the initial incident occurred.Crime report data is provided for Louisville Metro Police Divisions only; crime data does not include smaller class cities, unless LMPD becomes involved in smaller agency incident.The data provided in this dataset is preliminary in nature and may have not been investigated by a detective at the time of download. The data is therefore subject to change after a complete investigation. This data represents only calls for police service where a police incident report was taken. Due to the variations in local laws and ordinances involving crimes across the nation, whether another agency utilizes Uniform Crime Report (UCR) or National Incident Based Reporting System (NIBRS) guidelines, and the results learned after an official investigation, comparisons should not be made between the statistics generated with this dataset to any other official police reports. Totals in the database may vary considerably from official totals following the investigation and final categorization of a crime. Therefore, the data should not be used for comparisons with Uniform Crime Report or other summary statistics.Contact:Ivan Benitez, Ph.D.Gun Violence Data FellowOffice for Safe and Healthy Neighborhoodsivan.benitez@louisvilleky.gov

Weapons Sales in Australia decreased to 75 SIPRI TIV Million in 2024 from 80 SIPRI TIV Million in 2023. Australia Weapons Sales - values, historical data, forecasts and news - updated on September of 2025.

Here are a few use cases for this project:

1. Security Surveillance: This model could be used in video surveillance systems to detect unusual or potentially threatening gestures, such as someone carrying a gun. Real-time alerts could be sent to security staff, triggering a rapid response.

2. Content Moderation: Social media platforms or other user content sharing platforms could use this model to automatize the process of content filtering, identifying potential violent or harmful imagery that involves people with guns, ensuring a safe digital environment for users.

3. Video Game Development: It could be used to enhance realism in video games or simulations by identifying characters in different scenarios or postures, for example, distinguishing between armed and unarmed characters.

4. Forensic Investigations: Law enforcement agencies could use this model to identify persons of interest in video evidence, especially focusing on distinguishing armed individuals in crime-related scenarios.

5. Traffic Control Systems: This model might be used in traffic monitoring to detect and report possible violations or threats, such as cases where someone displays a firearm in a vehicle or public place. A real-time alert system could be built upon this that informs the local law enforcement agencies.

THIS DATASET WAS LAST UPDATED AT 8:11 PM EASTERN ON AUG. 30

OVERVIEW

2019 had the most mass killings since at least the 1970s, according to the Associated Press/USA TODAY/Northeastern University Mass Killings Database.

In all, there were 45 mass killings, defined as when four or more people are killed excluding the perpetrator. Of those, 33 were mass shootings . This summer was especially violent, with three high-profile public mass shootings occurring in the span of just four weeks, leaving 38 killed and 66 injured.

A total of 229 people died in mass killings in 2019.

The AP's analysis found that more than 50% of the incidents were family annihilations, which is similar to prior years. Although they are far less common, the 9 public mass shootings during the year were the most deadly type of mass murder, resulting in 73 people's deaths, not including the assailants.

One-third of the offenders died at the scene of the killing or soon after, half from suicides.

About this Dataset

The Associated Press/USA TODAY/Northeastern University Mass Killings database tracks all U.S. homicides since 2006 involving four or more people killed (not including the offender) over a short period of time (24 hours) regardless of weapon, location, victim-offender relationship or motive. The database includes information on these and other characteristics concerning the incidents, offenders, and victims.

The AP/USA TODAY/Northeastern database represents the most complete tracking of mass murders by the above definition currently available. Other efforts, such as the Gun Violence Archive or Everytown for Gun Safety may include events that do not meet our criteria, but a review of these sites and others indicates that this database contains every event that matches the definition, including some not tracked by other organizations.

This data will be updated periodically and can be used as an ongoing resource to help cover these events.

Using this Dataset

To get basic counts of incidents of mass killings and mass shootings by year nationwide, use these queries:

Mass killings by year

Mass shootings by year

To get these counts just for your state:

Filter killings by state

Definition of "mass murder"

Mass murder is defined as the intentional killing of four or more victims by any means within a 24-hour period, excluding the deaths of unborn children and the offender(s). The standard of four or more dead was initially set by the FBI.

This definition does not exclude cases based on method (e.g., shootings only), type or motivation (e.g., public only), victim-offender relationship (e.g., strangers only), or number of locations (e.g., one). The time frame of 24 hours was chosen to eliminate conflation with spree killers, who kill multiple victims in quick succession in different locations or incidents, and to satisfy the traditional requirement of occurring in a “single incident.”

Offenders who commit mass murder during a spree (before or after committing additional homicides) are included in the database, and all victims within seven days of the mass murder are included in the victim count. Negligent homicides related to driving under the influence or accidental fires are excluded due to the lack of offender intent. Only incidents occurring within the 50 states and Washington D.C. are considered.

Methodology

Project researchers first identified potential incidents using the Federal Bureau of Investigation’s Supplementary Homicide Reports (SHR). Homicide incidents in the SHR were flagged as potential mass murder cases if four or more victims were reported on the same record, and the type of death was murder or non-negligent manslaughter.

Cases were subsequently verified utilizing media accounts, court documents, academic journal articles, books, and local law enforcement records obtained through Freedom of Information Act (FOIA) requests. Each data point was corroborated by multiple sources, which were compiled into a single document to assess the quality of information.

In case(s) of contradiction among sources, official law enforcement or court records were used, when available, followed by the most recent media or academic source.

Case information was subsequently compared with every other known mass murder database to ensure reliability and validity. Incidents listed in the SHR that could not be independently verified were excluded from the database.

Project researchers also conducted extensive searches for incidents not reported in the SHR during the time period, utilizing internet search engines, Lexis-Nexis, and Newspapers.com. Search terms include: [number] dead, [number] killed, [number] slain, [number] murdered, [number] homicide, mass murder, mass shooting, massacre, rampage, family killing, familicide, and arson murder. Offender, victim, and location names were also directly searched when available.

This project started at USA TODAY in 2012.

Contacts

Contact AP Data Editor Justin Myers with questions, suggestions or comments about this dataset at jmyers@ap.org. The Northeastern University researcher working with AP and USA TODAY is Professor James Alan Fox, who can be reached at j.fox@northeastern.edu or 617-416-4400.

Firearms background checks for the USA for 2012 (Jan-Nov) and since 1999.These statistics represent the number of firearm background checks initiated through the NICS. They do not represent the number of firearms sold. NICS is used by Federal Firearms Licensees (FFLs) to instantly determine whether a prospective buyer is eligible to buy firearms or explosives. Before ringing up the sale, cashiers call in a check to the FBI or to other designated agencies to ensure that each customer does not have a criminal record or isn't otherwise ineligible to make a purchase. More than 100 million such checks have been made in the last decade, leading to more than 700,000 denials. More information on NICS - http://www.fbi.gov/about-us/cjis/nics Some really useful informations such as the rate of checks per 1000 people. All data is provided by state. Downloaded from the Guardian Datablog - http://www.guardian.co.uk/news/datablog/2012/dec/17/how-many-guns-us and then joined to USA States data http://geocommons.com/overlays/21424. Gun data originally from FBI http://www.fbi.gov/about-us/cjis/nics. GIS vector data. This dataset was first accessioned in the EDINA ShareGeo Open repository on 2012-12-17 and migrated to Edinburgh DataShare on 2017-02-21.

Detecting People With Guns_2

## Overview

Detecting People With Guns_2 is a dataset for object detection tasks - it contains People Weapons annotations for 2,620 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).

The Firearms Records Bureau publishes data on firearms licensing by licensing authority and firearms transactions by gun dealers.