Notable for labeled images of UNO playing cards specifically curated for object detection tasks within the RF100 benchmark. It provides bounding box annotations for identifying various card types, colors, and symbols found in a standard UNO deck.
Use Cases
- Train a YOLO or SSD model to detect card positions using the provided bounding box coordinates.
- Classify card colors and values by mapping image regions to the specific class labels in the dataset.
- Benchmark the performance of object detection architectures against other datasets in the RF100 collection.
Strengths
- Includes bounding box annotations for individual UNO cards.
- Part of the Roboflow 100 (RF100) benchmark for evaluating object detection models.
- Features labels for distinct card types including numbers and action symbols.
- Standardized for use in computer vision pipelines targeting small-to-medium object recognition.