FIgLib & SmokeyNet: Dataset and Deep Learning Model for Real-Time Wildland Fire Smoke Detection

Image | Wildfire | Preparedness

FIgLib is a dataset for real-time wildfire smoke detection. It contains 24,800 labeled wildfire smoke images of Southern California and is introduced for binary classification.

  • ML task type: Binary classification
  • Data Source: Earth Observation Data and GeoSpatial Imagery (Fixed-view cameras)
  • Size: 24,800 Images
  • Timespan: June 2016 - July 2021
  • Geographical Coverage: Southern California
  • Baseline Information
  • Evaluated on: SmokeyNet, MobileNetV3Large, MobileNet with Feature Pyramid Network, EfficientNet-B0, Data Efficient Image Transformer (DeiT-Tiny)
  • Metrics used: Accuracy, Precision, Recall, F1-Score
  • Results as reported in original paper: Accuracy: 83.62, Precision: 90.85, Recall: 76.11, F-Score: 82.83

Anshuman Dewangan, Yash Pande, Hans-Werner Braun, Frank Vernon, Ismael Perez, Ilkay Altintas, Garrison W Cottrell, and Mai H Nguyen. Figlib & smokeynet: Dataset and deep learning model for real-time wildland fire smoke detection. Remote Sensing, 14(4):1007, 2022.