Datasets:
Update dataset card metadata and paper links
#2
by nielsr HF Staff - opened
README.md
CHANGED
|
@@ -1,52 +1,53 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
dataset_info:
|
| 3 |
features:
|
| 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 |
splits:
|
| 29 |
-
|
| 30 |
-
task_categories:
|
| 31 |
-
- image-classification
|
| 32 |
-
- image-to-image
|
| 33 |
-
language:
|
| 34 |
-
- en
|
| 35 |
-
license: cc-by-4.0
|
| 36 |
tags:
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
pretty_name: "NearID (Multi-View Identity Dataset)"
|
| 43 |
-
size_categories:
|
| 44 |
-
- 10K<n<100K
|
| 45 |
---
|
| 46 |
|
| 47 |
# NearID — Multi-View Identity Dataset
|
| 48 |
|
| 49 |
-
[](https://huggingface.co/Aleksandar/nearid-siglip2) [ project. Each sample contains multiple views of the **same identity** rendered in different backgrounds/contexts.
|
| 52 |
|
|
@@ -95,7 +96,7 @@ neg = load_dataset("Aleksandar/NearID-Flux")
|
|
| 95 |
## Related
|
| 96 |
|
| 97 |
- **Model:** [Aleksandar/nearid-siglip2](https://huggingface.co/Aleksandar/nearid-siglip2) — NearID identity embedding model
|
| 98 |
-
- **Paper:** [NearID: Identity Representation Learning via Near-identity Distractors](https://
|
| 99 |
- **Code:** [github.com/Gorluxor/NearID](https://github.com/Gorluxor/NearID)
|
| 100 |
|
| 101 |
## License & Attribution
|
|
@@ -109,4 +110,4 @@ This dataset is released under [CC-BY-4.0](https://creativecommons.org/licenses/
|
|
| 109 |
title={NearID: Identity Representation Learning via Near-identity Distractors},
|
| 110 |
author={Cvejic, Aleksandar and Abdal, Rameen and Eldesokey, Abdelrahman and Ghanem, Bernard and Wonka, Peter}
|
| 111 |
}
|
| 112 |
-
```
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
license: cc-by-4.0
|
| 5 |
+
size_categories:
|
| 6 |
+
- 10K<n<100K
|
| 7 |
+
task_categories:
|
| 8 |
+
- image-feature-extraction
|
| 9 |
+
- image-classification
|
| 10 |
+
- image-to-image
|
| 11 |
+
pretty_name: NearID (Multi-View Identity Dataset)
|
| 12 |
dataset_info:
|
| 13 |
features:
|
| 14 |
+
- name: id
|
| 15 |
+
dtype: int64
|
| 16 |
+
- name: category
|
| 17 |
+
dtype: string
|
| 18 |
+
- name: category_description
|
| 19 |
+
dtype: string
|
| 20 |
+
- name: img1
|
| 21 |
+
dtype: image
|
| 22 |
+
- name: img2
|
| 23 |
+
dtype: image
|
| 24 |
+
- name: img3
|
| 25 |
+
dtype: image
|
| 26 |
+
- name: n_images
|
| 27 |
+
dtype: int64
|
| 28 |
+
- name: objaverse_id
|
| 29 |
+
dtype: string
|
| 30 |
+
- name: prompts1
|
| 31 |
+
dtype: string
|
| 32 |
+
- name: prompts2
|
| 33 |
+
dtype: string
|
| 34 |
+
- name: prompts3
|
| 35 |
+
dtype: string
|
| 36 |
+
- name: quality
|
| 37 |
+
dtype: string
|
| 38 |
splits:
|
| 39 |
+
- name: train
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
tags:
|
| 41 |
+
- nearid
|
| 42 |
+
- identity-embedding
|
| 43 |
+
- multi-view
|
| 44 |
+
- synthetic
|
| 45 |
+
- metric-learning
|
|
|
|
|
|
|
|
|
|
| 46 |
---
|
| 47 |
|
| 48 |
# NearID — Multi-View Identity Dataset
|
| 49 |
|
| 50 |
+
[](https://huggingface.co/Aleksandar/nearid-siglip2) [](https://huggingface.co/papers/2604.01973) [](https://gorluxor.github.io/NearID/) [](https://github.com/Gorluxor/NearID) [](https://www.kaust.edu.sa/) [](https://research.snap.com/)
|
| 51 |
|
| 52 |
This is the **base positives dataset** for the [NearID](https://huggingface.co/Aleksandar/nearid-siglip2) project. Each sample contains multiple views of the **same identity** rendered in different backgrounds/contexts.
|
| 53 |
|
|
|
|
| 96 |
## Related
|
| 97 |
|
| 98 |
- **Model:** [Aleksandar/nearid-siglip2](https://huggingface.co/Aleksandar/nearid-siglip2) — NearID identity embedding model
|
| 99 |
+
- **Paper:** [NearID: Identity Representation Learning via Near-identity Distractors](https://huggingface.co/papers/2604.01973)
|
| 100 |
- **Code:** [github.com/Gorluxor/NearID](https://github.com/Gorluxor/NearID)
|
| 101 |
|
| 102 |
## License & Attribution
|
|
|
|
| 110 |
title={NearID: Identity Representation Learning via Near-identity Distractors},
|
| 111 |
author={Cvejic, Aleksandar and Abdal, Rameen and Eldesokey, Abdelrahman and Ghanem, Bernard and Wonka, Peter}
|
| 112 |
}
|
| 113 |
+
```
|