AOW-YOLO: Model Architecture Details for Smoking Detection

Ruishi Liang published details on the AOW-YOLO model architecture on 2026-05-06 via figshare. The 5.5 KB XLS file describes a novel object detection model designed to identify smoking behavior in complex construction site environments. The model reportedly outperforms existing lightweight models in mAP metrics and achieves faster inference speeds.

Use Cases

• Implementing the AOW-YOLO architecture based on the described Adaptive Occlusion and Weighting IoU (AOWIoU) loss function.
• Developing construction site safety monitoring systems based on the described model for detecting small objects in cluttered scenes.
• Benchmarking lightweight detection models based on the reported mAP50 and mAP50:95 performance metrics.
• Studying the integration of the Spatial Grouped-Pointwise (SGP) convolution module for cross-group information fusion.

Strengths

• Model performance is quantified with specific mAP50 and mAP50:95 metrics.
• Inference speed improvement of roughly 31.6% over a baseline model (YOLO11n) is reported.
• The dataset is openly licensed under CC-BY-4.0.

Limitations

• The dataset is very small (5.5 KB), suggesting it contains only model architecture details, not training or evaluation data.
• Row count and column-level documentation are absent; the actual data structure must be inferred after download.
• The description focuses on the model, not the accompanying data file's specific contents.

Provenance

Source

Ruishi Liang via figshare.

Collection Method

Likely contains tabular data detailing model architecture, components, or performance metrics.

Freshness

Last updated 2026-05-06 17:45:40.