slarabi/cognitive-remediation

Cognitive Remediation Dataset

Overview

The Cognitive Remediation Dataset is a collection of videos illustrating cognitive remediation exercises. It was created to support research in artificial intelligence, computer vision, and healthcare. The developed dataset is fundamentally built upon the physical interactions between the patient and the physical objects within the precisely defined miniature urban scene. Rather than relying on simulated environments, this dataset captures real-world, hands-on manipulations of miniature vehicles navigating through the custom-built infrastructure. It comprises a wide variety of clinical and behavioral actions specifically designed to reflect both correct and incorrect execution of cognitive exercises, providing a robust foundation for subsequent VLM training and evaluation.

Dataset Contents

The collected data is rigorously organized into a hierarchical architecture consisting of 5 main categories and 17 specific actions.

Directory Parking

Actions: - Parking correctly (within lines) - Parking incorrectly (random placement outside designated spaces) - Parking incorrectly (not aligned with lines, occupying two spaces) - Parking on the sidewalk

Directory ParkingEntranceExit

Actions: - Entering through the entrance - Exiting through the exit - Entering through the exit (wrong way) - Exiting through the entrance (wrong way)

Directory Road

Actions: - Driving in the correct direction of the road - Driving in the wrong direction of the road

Directory Roundabout

Actions: - Turning around the roundabout correctly (correct direction, full loop) - Turning around the roundabout in the wrong direction - Completing only part of the roundabout - Crossing the roundabout directly

Directory Sidewalk

Actions: - Driving on the sidewalk

Applications

This dataset can be used for:

• Deep learning
• Computer vision
• Human-computer interaction
• Cognitive rehabilitation research

License

license: cc-by-nc-nd-4.0

Citation

If you use this dataset, please cite:

@dataset{larabi2025cognitiveremediation,
  author = {Ait Mehdi Nassira, Temam Milissa, Slimane Larabi},
  title = {Cognitive Remediation Dataset},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/slarabi/cognitive-remediation}
}