Full Dataset Object Detection Dataset and Pre-Trained Model by ENGS 89

Images

24858 images

Explore Dataset

Overview This dataset is for the Hanover Bike Walk Census Tool YOLO model. It contains pedestrians, people riding on bicycles, people riding on e-scooters, and people in wheelchairs. We have trained a custom YOLOv8s model on this dataset and have found that the model generalizes quite well to stationary videos of urban environments. If you want to see the full project, please visit this github: https://github.com/rolson24/BWCT-Tracker/tree/electron-app

Classes

Pedestrian
Bikex
Scooter
Wheelchair

Detailed Annotation Style We decided to annotate the pedestrians as full people including their head, feet, and their arms. If a pedestrian's arms are sticking directly out to the side, we choose to not include them in the bounding box to make the box smaller for use with an object tracker. For bikes, scooters, and wheelchairs we included both the device that people are using for movement as well as the person's full body. This decision was made to ensure that the other classes would not be incorrectly detected as a pedestrian on top of an object.

Trained Model API

This project has a trained model available that you can try in your browser and use to get predictions via our Hosted Inference API and other deployment methods.

Cite This Project

If you use this dataset in a research paper, please cite it using the following BibTeX:

@misc{
                            full-dataset-ncrgq_dataset,
                            title = { Full Dataset Dataset },
                            type = { Open Source Dataset },
                            author = { ENGS 89 },
                            howpublished = { \url{ https://universe.roboflow.com/engs-89/full-dataset-ncrgq } },
                            url = { https://universe.roboflow.com/engs-89/full-dataset-ncrgq },
                            journal = { Roboflow Universe },
                            publisher = { Roboflow },
                            year = { 2024 },
                            month = { mar },
                            note = { visited on 2024-05-02 },
                            }