Skip to content
work

Cigarette Brand Detection Model

Object detection model for distinguishing cigarette brands using computer vision.

Cigarette Brand Detection Model, image 1 of 4

Solution

  • Multi-brand cigarette detection
  • Custom dataset creation and labeling
  • Detectron model integration
  • Bounding box annotation system
  • Optimized detection accuracy

My contribution

  • Built sophisticated object detection model capable of distinguishing different cigarette brands.
  • Utilized pre-trained Detectron model as foundation for custom brand detection functionality.
  • Collected and curated comprehensive dataset of cigarette box images for training.
  • Applied bounding box annotations and labeling to identify cigarette box regions accurately.
  • Optimized model configuration to achieve best possible detection accuracy.