anywaylabs/synthetic-driver-monitoring-detection

Synthetic DMS – Driver Monitoring System Dataset by AnywayLabs.ai

Need a custom synthetic dataset for your own road safety detection use case?

This dataset is an open-source sample of our synthetic data generation work at AnywayLabs.

If you're working on:

• industrial defect detection
• visual inspection
• supervised anomaly detection
• hard-to-collect defect classes
• synthetic data for computer vision training

You can request a custom synthetic dataset here, or email: contact@anywaylabs.ai

Dataset Summary

This dataset contains fully synthetic images for driver monitoring and distraction detection, generated by AnywayLabs.ai.

The dataset is designed for supervised object detection of driver behavior, covering four distraction classes across multiple camera viewpoints (FOV). It targets real-world in-cabin monitoring scenarios and is intended to train models that detect risky driver behaviors in automotive environments.

Key Features

• Fully synthetic driver monitoring dataset
• Supervised object detection formulation
• YOLO-format annotations
• Multiple camera fields-of-view (FOV1, FOV2, FOV3 and variants)
• 4 distraction behavior classes
• 1,356 annotated training images

---

Sample Images

FOV1	FOV2	FOV3

Dataset Structure

Format

• Task: Object Detection (Driver Behavior / Distraction Detection)
• Annotation Format: YOLO (normalized bounding boxes)
• Image Resolution: 1376 × 768

Directory Structure

images/
  train/
labels/
  train/
data.yaml
metadata_train.jsonl

---

Data Generation

The dataset is generated using AnywayLabs.ai's proprietary synthetic dataset generation framework.

Capabilities

• Behavior-level control:

  • Action type (e.g., drinking, yawning, calling, texting)
  • Pose variation and body positioning
  • Occlusion and partial visibility

• Scene-level control:

  • Lighting conditions
  • Interior background variation
  • Camera placement and field-of-view

• Structural consistency:

  • Driver body geometry preserved
  • Realistic action integration

Design Philosophy

Instead of replicating real data exactly, the dataset is designed to:

• Expand the behavioral distribution beyond limited real samples
• Introduce controlled variation across action appearance
• Improve robustness to unseen real-world driver behaviors

---

Annotation Process

• Bounding boxes are generated for each behavior instance

• Labels follow YOLO format:

  class_id x_center y_center width height
  

• Bounding box tightness is manually reviewed

• Annotation consistency is maintained across all camera views

---

Limitations

• Validation sets are not included in the public release
• Performance may vary across camera placements and lighting conditions
• Bounding boxes do not capture fine-grained body pose (compared to keypoint annotations)

---

Usage

Recommended Use Cases

• Driver distraction detection model training
• Synthetic pretraining before fine-tuning on real in-cabin data
• Automotive safety system development
• Multi-view behavior recognition research

Compatible Frameworks

• Ultralytics YOLO (recommended)
• Detectron2 (after conversion)
• Any YOLO-compatible pipeline

---

Comparison with Real-World DMS Datasets

Feature	Real-World DMS Datasets	This Dataset
Data Source	Real vehicle recordings	Fully synthetic
Privacy Concerns	High (driver identity)	None
Annotation Effort	Manual, costly	Automated
Behavioral Coverage	Limited by recording	Configurable
Class Balance	Often imbalanced	Controllable
Training Paradigm	Supervised	Fully supervised

---

Need a custom synthetic dataset for your own road safety detection use case?

This dataset is an open-source sample of our synthetic data generation work at AnywayLabs.

If you're working on:

• industrial defect detection
• visual inspection
• supervised anomaly detection
• hard-to-collect defect classes
• synthetic data for computer vision training

You can request a custom synthetic dataset here, or email: contact@anywaylabs.ai

Citation

If you use this dataset, please cite:

Synthetic DMS Driver Monitoring Dataset – AnywayLabs.ai