Datasets:

Voxel51
/

SynthHuman

Modalities:

Formats:

Languages:

Size:

ArXiv:

Tags:

Libraries:

License:

Dataset card Data Studio Files Files and versions

harpreetsahota commited on Aug 21, 2025

Commit

d932806

·

verified ·

1 Parent(s): 0550fe2

Update README.md

Files changed (1) hide show

README.md +3 -4

README.md CHANGED Viewed

@@ -57,10 +57,7 @@ license: cdla-permissive-2.0
 # Dataset Card for SynthHuman
-<!-- Provide a quick summary of the dataset. -->
 This is a [FiftyOne](https://github.com/voxel51/fiftyone) dataset with 3000 samples.
@@ -112,7 +109,9 @@ Unlike scan-based synthetic datasets, SynthHuman uses high-fidelity procedural g
 - **Paper:** DAViD: Data-efficient and Accurate Vision Models from Synthetic Data (arXiv:2507.15365)
 - **Parsing to FiftyOne format:** https://github.com/harpreetsahota204/synthhuman_to_fiftyone
 ## Uses

 # Dataset Card for SynthHuman
+![image/png](synthhuman-mq.gif)
 This is a [FiftyOne](https://github.com/voxel51/fiftyone) dataset with 3000 samples.
 - **Paper:** DAViD: Data-efficient and Accurate Vision Models from Synthetic Data (arXiv:2507.15365)
+#### Additional resources
 - **Parsing to FiftyOne format:** https://github.com/harpreetsahota204/synthhuman_to_fiftyone
+- **Checkout this notebook to learn how to do some cool analysis with this dataset**:
 ## Uses