Description:

This dataset was created to develop a robust machine learning model capable of differentiating between knives and pistols in images. The primary goal is to aid in object recognition tasks, particularly in security applications where identifying potential weapons is crucial.

Download Dataset

Dataset Structure

The dataset is divided into four key folders, each containing images meant for both training and evaluation of machine learning models:

Knife: Contains images specifically focused on knives. These are used to train the knife recognition model.

eval_Knife: Designed for evaluating the knife detection model’s accuracy and its ability to make reliable predictions.

Pistol: Contains images of pistols, design to train the model in distinguishing pistols from other objects.

eval_Pistol: Use to test and evaluate the pistol detection model, ensuring that it can effectively predict pistol-related outcomes.

Additional Features

Image Variations: The dataset includes various angles, lighting conditions, and backgrounds to ensure robustness in diverse real-world scenarios.

Data Augmentation: To improve model generalization, data augmentation techniques such as rotation, scaling, and cropping can be apply to simulate different environments.

Annotation Files: The dataset includes label annotations, providing bounding boxes around the objects (knife or pistol) within each image, facilitating precise object localization tasks.

Use Case Examples: This dataset is particularly suited for applications in airport security, automate surveillance systems, and law enforcement technologies where accurate weapon detection is critical.

This dataset is sourced from Kaggle.