This dataset was curated and annotated by Ishaan Singh, a high school student from India.
The original dataset (v1) is composed of 166 images of various cars present in a junkyard. Training Set: 116 images, Validation Set: 33 images, Testing Set: 17 images.
The dataset is available under the Public License.
Ishaan ultimately used this dataset to create a "Drone Surveillance" system to count the cars using YOLOv5 & Deep SORT (Simple Online and Realtime Tracking with a Deep Association Metric) for a contest organized by ComputerVisionZone.
Here is a video of his final submission for the contest:
Video of Ishaan's Final Model
You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.
Version 1 (v1) - 166 images
- Preprocessing: Auto-Orient and Resize (Stretch to 416x416)
- Augmentations: No augmentations applied
- Training Metrics: This version of the dataset was not trained