Vape Related Social Media Object Detection

This is the code for object detection. this codebase has a similar version at https://github.com/tanviaanand/Vape-Detection

The code in this repository was used to develop a computer vision model to identify e-cigarette content in TikTok videos. Specifically, object detection techniques were used to detect e-cigarette-specific images by analyzing screenshots captured from videos with e-cigarette-related hashtags. Image annotation data was created using bounding boxes to indicate the presence of e-cigarettes, whether alone, in use, or accompanied by smoke or vapor. The YoloV7 model was employed for object detection, to classify e-cigarette content (including vape devices and challenging entities such as vapor clouds and individuals using e-cigarettes).