Hybrid Vehicle Logo Dataset with Synthetic Training Images

A dataset for vehicle brand logo classification and identification from surveillance and automated number plate recognition feeds. It contains real-world images from Kozhikode District, Kerala, India, and a training set augmented with 1,170 high-fidelity synthetic images. The dataset was created by Arun Natarajan and last updated on 2026-05-24.

Use Cases

• Train logo detection models based on extreme close-up, cropped bounding box viewframes of car grilles and trunk rear panels.
• Improve model robustness for rare vehicle models using a hybrid composition of real and synthetic data.
• Test model performance in adverse conditions using synthetic images simulating extreme low-light, diverse camera angles, and adverse weather.
• Develop ANPR and surveillance systems for brand identification based on field-captured frames from urban traffic junctions.

Strengths

• Includes 1,170 high-fidelity synthetic images generated via Midjourney to expand classification boundaries for minority classes.
• Real-world component sourced from high-angled urban surveillance and ANPR systems across multiple primary junctions.
• Data is structured in the standard YOLO Object Detection format for seamless integration with deep learning frameworks like YOLOv8.
• All data instances were manually bounded and verified for high precision using the CVAT annotation tool.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data may reflect geographic bias inherent to its collection in Kozhikode District, Kerala, India.

Provenance

Source

Arun Natarajan, with real-world data captured from traffic surveillance streams in Kozhikode District, Kerala, India.

Collection Method

Real-world keyframes captured via surveillance/ANPR systems; synthetic images generated via Midjourney; annotations done manually with CVAT.

Freshness

Last updated 2026-05-24 16:40:37; freshness should be verified.

Geography

Kozhikode District, Kerala, India.