File size: 2,210 Bytes
4a617e6 7b35a01 9c1d2a1 9d234b4 9c1d2a1 9d234b4 9c1d2a1 50b9bd8 9c1d2a1 50b9bd8 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 | ---
license: cc-by-sa-4.0
tags:
- object_detection
- uncertainty
- soft_labels
- label_errors
pretty_name: Refined Annotations with Soft Labels for Popular Object Detection Datasets
---
Refined Annotations with Soft Labels for Popular Object Detection Datasets
===============================
These datasets contain refined annotations for four popular object detection benchmark datasets:
COCO 2017 validation data, Pascal VOC 2012 detection data for both training and validation split, KITTI 2D object detection training data, and Cityscapes training and validation data. Each annotation consists of a bounding box (xmin, ymin, xmax, ymax) and a soft label annotation ("postprocessed_softlabel") as well as original annotator responses for object categories by annotator ID, including answer times.
Cityscapes Annotations are available through the official Cityscapes platform (https://www.cityscapes-dataset.com/downloads/, "gtBboxValidated_trainval.zip (3.2MB) [md5]")
See the accompanying publication and GitHub repo for more details:
Paper: https://arxiv.org/abs/2609.21822
GitHub: https://github.com/JonathanKlees/Uncertainty-aware-Object-Detection
If you use these annotations, please cite:
```bibtex
@inproceedings{penquitt2026object,
title = {{Object Detection Benchmarks are Incomplete: The Role of Label Errors and Annotation Uncertainty}},
author = {Penquitt, Sarina and Klees, Jonathan and van Betteray, Antonia and Jashnieh, Parssa and Stehr, Peter and Rottmann, Matthias and Schmarje, Lars},
booktitle = {Advances in Neural Information Processing Systems Evaluations \& Datasets Track},
year = {2026},
}
```
License for Dataset Annotations
===============================
Copyright (c) 2026 Sarina Penquitt, Jonathan Klees, Antonia van Betteray,
Parssa Jashnieh, Peter Stehr, Matthias Rottmann, Lars Schmarje
The annotations created as part of our annotation pipeline are licensed
under the Creative Commons Attribution-ShareAlike 4.0 International License
(CC BY-SA 4.0).
To view a copy of this license, visit:
https://creativecommons.org/licenses/by-sa/4.0/
Please note that this license only applies to the annotations provided here and original datasets have their own licensing. |