Sign inGet started
Dataset ยท Hugging Face

haydarkadioglu/brand-eye-dataset

๐Ÿš— Brand Eye Dataset - Automotive Brand Detection A comprehensive dataset for automotive brand/logo detection and recognition, formatted for YOLO training. This dataset contains high-quality images with precise annotations for 33 different automotive brands. ๐Ÿ“Š Dataset Overview This dataset is specifically designed for object detection tasks focusing on automotive brand recognition in various contexts including vehicles, logos, and brand emblems. ๐Ÿ“ˆโ€ฆ See the full description on the dataset page: https://huggingface.co/datasets/haydarkadioglu/brand-eye-dataset.

Author / organizationhaydarkadioglu
License (as declared)apache-2.0
Downloads1.1K (as reported by Hugging Face)
Likes5
Created23 September 2025
Last updated at source24 September 2025
computer-visionobject-detectionyolobrand-detectionlogo-detectioncar-brands
Embed this entry
<iframe src="https://silytics.si/embed/item/e78450ff-92f9-4760-8d1f-6852ede228d7" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 3 days

1,073 now (+64 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

Recent changes

  • 16 Oct 2026 Description: ๐Ÿš— Brand Eye Dataset - Automotive Brand Detection A comprehensive dataset for automotive brand/logo detection and recognition, formatted for YOLO training. This dataset contains high-quality images with precise annotations for 33 different โ†’ ๐Ÿš— Brand Eye Dataset - Automotive Brand Detection A comprehensive dataset for automotive brand/logo detection and recognition, formatted for YOLO training. This dataset contains high-quality images with precise annotations for 33 different

Changes detected by Silytics between refreshes of the source.

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.