YOLOv1 Object Detection Dataset

YOLOv1 is a foundational dataset for training and evaluating the 'You Only Look Once' real-time object detection model. The dataset contains images annotated with bounding boxes and class labels for multiple objects. It was introduced by Joseph Redmon et al. in the 2015 paper 'You Only Look Once: Unified, Real-Time Object Detection'.

Use Cases

• Train a model to predict bounding box coordinates and class probabilities from input images using the YOLO architecture.
• Benchmark object detection speed and mean average precision (mAP) metrics against the original YOLOv1 paper results.
• Use the provided image annotations to study the model's performance on specific object classes like 'person', 'car', or 'dog'.
• Analyze the trade-off between detection accuracy and inference speed inherent in the unified detection approach.

Strengths

• Serves as the benchmark dataset for the influential YOLOv1 model, cited thousands of times in computer vision literature.
• Provides annotations for multiple objects per image, enabling multi-object detection training.

Limitations

• The original YOLOv1 model and dataset are outdated compared to modern versions (YOLOv5-v10) with higher accuracy and more features.
• Specific dataset statistics like the number of images, classes, and annotation counts are not provided in this metadata.

Provenance

Source

Joseph Redmon, Santosh Divvala, Ross Girshick, Ali Farhadi

Collection Method

Created for the research and training of the YOLOv1 neural network model.