Datasets:
Add task category and link to paper
Browse filesHi! I'm Niels from the Hugging Face community science team.
I've opened this pull request to improve the dataset card for your BDD100K training split. This PR adds the `image-segmentation` task category to the YAML metadata, which helps researchers find your dataset when filtering by task. I've also ensured the paper is correctly linked to help with visibility and discoverability.
Thanks for sharing your work with the community!
README.md
CHANGED
|
@@ -1,9 +1,11 @@
|
|
| 1 |
---
|
| 2 |
license: other
|
| 3 |
-
license_name: bdd100k-license
|
| 4 |
-
license_link: LICENSE
|
| 5 |
source_datasets:
|
| 6 |
- bdd100k
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
dataset_info:
|
| 8 |
features:
|
| 9 |
- name: very_lossy_0
|
|
@@ -111,7 +113,7 @@ configs:
|
|
| 111 |
|
| 112 |
# BDD100K Train
|
| 113 |
|
| 114 |
-
This dataset was created for [
|
| 115 |
|
| 116 |
The underlying data is derived from the [BDD100K](https://www.bdd100k.com/) driving video dataset. It contains 69,800 training sequences of 15 frames each.
|
| 117 |
|
|
@@ -149,7 +151,7 @@ This dataset is distributed under the BDD100K license:
|
|
| 149 |
```bibtex
|
| 150 |
@inproceedings{jacobellis2026dedelayed,
|
| 151 |
title = {Dedelayed: Deleting Remote Inference Delay via On-Device Correction},
|
| 152 |
-
author = {Jacobellis, Dan and Ulhaq, Mateen and Racap{\'e}, Fabien and Choi, Hyomin and Yadwadkar, Neeraja J.},
|
| 153 |
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
|
| 154 |
year = {2026},
|
| 155 |
note = {To appear}
|
|
@@ -169,4 +171,4 @@ This dataset is distributed under the BDD100K license:
|
|
| 169 |
booktitle={The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
|
| 170 |
year={2017}
|
| 171 |
}
|
| 172 |
-
```
|
|
|
|
| 1 |
---
|
| 2 |
license: other
|
|
|
|
|
|
|
| 3 |
source_datasets:
|
| 4 |
- bdd100k
|
| 5 |
+
license_name: bdd100k-license
|
| 6 |
+
license_link: LICENSE
|
| 7 |
+
task_categories:
|
| 8 |
+
- image-segmentation
|
| 9 |
dataset_info:
|
| 10 |
features:
|
| 11 |
- name: very_lossy_0
|
|
|
|
| 113 |
|
| 114 |
# BDD100K Train
|
| 115 |
|
| 116 |
+
This dataset was created for [DeDelayed: Deleting Remote Inference Delay via On-Device Correction](https://huggingface.co/papers/2510.13714) (CVPR 2026). Code is available at [InterDigitalInc/dedelayed](https://github.com/InterDigitalInc/dedelayed).
|
| 117 |
|
| 118 |
The underlying data is derived from the [BDD100K](https://www.bdd100k.com/) driving video dataset. It contains 69,800 training sequences of 15 frames each.
|
| 119 |
|
|
|
|
| 151 |
```bibtex
|
| 152 |
@inproceedings{jacobellis2026dedelayed,
|
| 153 |
title = {Dedelayed: Deleting Remote Inference Delay via On-Device Correction},
|
| 154 |
+
author = {Jacobellis, Dan and Ulhaq, Mateen and Racap{\\'e}, Fabien and Choi, Hyomin and Yadwadkar, Neeraja J.},
|
| 155 |
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
|
| 156 |
year = {2026},
|
| 157 |
note = {To appear}
|
|
|
|
| 171 |
booktitle={The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
|
| 172 |
year={2017}
|
| 173 |
}
|
| 174 |
+
```
|