datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
aria_synthetic_envs_mcmc_3dgs_newLicense Notice:This dataset is derived from the Aria Dataset.It follows the Aria Synthetic Environments Dataset License Agreement.See Aria License for details.
3dgsmatterport3d_region_mcmc_3dgsLicense Notice:This dataset is derived from Matterport3D.It follows the Matterport End User License Agreement for Academic Use of Model Data.See Matterport3D License for details.
scannet_mcmc_3dgs
Data Statistics
Scenes
Mean PSNR ↑
Mean SSIM ↑
Mean LPIPS ↓
Mean Depth L1 ↓
Mean #3DGS
Total #3DGS
1,613
30.17 dB
0.875
0.221
0.0151 m
1.000M
1.613B
License Notice:This dataset is derived from ScanNet and follows the ScanNet Terms of Use.See ScanNet Terms for details.
arkitscenes_mcmc_3dgs
Data Statistics
Scenes
Mean PSNR ↑
Mean SSIM ↑
Mean LPIPS ↓
Mean Depth L1 ↓
Mean #3DGS
Total #3DGS
1,290
31.63 dB
0.907
0.217
0.0051 m
1.149M
1.483B
scannetpp_v2_mcmc_3dgs
Data Statistics
Scenes
Mean PSNR ↑
Mean SSIM ↑
Mean LPIPS ↓
Mean Depth L1 ↓
Mean #3DGS
Total #3DGS
956
35.08 dB
0.959
0.062
0.0031 m
1.500M
1.434B
License Notice:This dataset is derived from ScanNet++ and follows the ScanNet++ Terms of Use.See ScanNet++ Terms for details.
3rscan_mcmc_3dgs
Data Statistics
Scenes
Mean PSNR ↑
Mean SSIM ↑
Mean LPIPS ↓
Mean Depth L1 ↓
Mean #3DGS
Total #3DGS
632
27.06 dB
0.874
0.344
0.0177 m
1.500M
948.000M
License Notice:This dataset is derived from 3RScan and follows the 3RScan Terms of Use.See 3RScan Terms for details.
dl3dv_10k_3dgs_newannotated-3DGS-artifacts
Puzzle Similarity
Project page | Paper | Code
This repository contains the dataset presented in the ICCV 2025 paper "Puzzle Similarity: A Perceptually-guided Cross-Reference Metric for Artifact Detection in 3D Scene Reconstructions"Authors: Nicolai Hermann, Jorge Condor, and Piotr Didyk
Dataset Description
The Dataset consists of 36 hand-selected 3D Gaussian Splatting renderings containing common reconstruction artefacts, (aligned) ground truths, human-annotated… See the full description on the dataset page: https://huggingface.co/datasets/nihermann/annotated-3DGS-artifacts.scannetpp_v1_mcmc_3dgs
Data Statistics
Scenes
Mean PSNR ↑
Mean SSIM ↑
Mean LPIPS ↓
Mean Depth L1 ↓
Mean #3DGS
Total #3DGS
330
28.89 dB
0.917
0.144
0.0201 m
1.552M
512.282M
License Notice:This dataset is derived from ScanNet++ and follows the ScanNet++ Terms of Use.See ScanNet++ Terms for details.
3DGS_Mesh_Envs
3DGS-Mesh Environments for HUGE-Bench
This dataset releases the 3DGS-Mesh digital twin environments for HUGE-Bench.
HUGE_Dataset_v0 contains the full trajectory dataset in LeRobot format, including train, test_seen, and test_unseen splits for all benchmark tasks. This repository contains the 3D assets required for rendering and collision-aware evaluation.
Each archive contains:
data_3d/<env_id>/3dgs_ply/point_cloud_utm50.ply
data_3d/<env_id>/terra_ply/simplified_mesh.obj… See the full description on the dataset page: https://huggingface.co/datasets/yu781986168/3DGS_Mesh_Envs.3DGS_latentsvehicle-3dgs
Vehicle-Class-Specific 3D Gaussian Splatting Dataset
🌐 Project Page | 📄 Paper | 🤗 Dataset | 💻 Code
A collection of 114 vehicle 3D Gaussian Splatting (3DGS) models in .ply format, each
corresponding to a distinct vehicle class (make / model / body type / year). The models were
produced by an automated multi-stage reconstruction pipeline from heterogeneous, multi-source
multi-view images. This is one of the first publicly available datasets of… See the full description on the dataset page: https://huggingface.co/datasets/muzafferardauslu/vehicle-3dgs.AnchorSplat-3DGS-SR-Train-15K
AnchorSplat 3DGS-SR Train 15K
This dataset release contains the 14,525 training scenes used for AnchorSplat 3DGS-SR training.
Each scene contains a low-resolution 3DGS input at 256 resolution and high-resolution supervision at 1024 resolution.
Files
parts/*.tar.zst: independent tar+zstd archives. Each part can be downloaded and extracted separately.
parts/*.tar.zst.sha256: per-part SHA256 checksum.
SHA256SUMS.txt: checksums for all archive parts.
MANIFEST.jsonl:… See the full description on the dataset page: https://huggingface.co/datasets/de233/AnchorSplat-3DGS-SR-Train-15K.3dgs-assetsmatterport3d_scene_mcmc_3dgs
Data Statistics
Scenes
Mean PSNR ↑
Mean SSIM ↑
Mean LPIPS ↓
Mean Depth L1 ↓
Mean #3DGS
Total #3DGS
90
29.04 dB
0.882
0.208
0.0161 m
12.000M
1.080B
License Notice:This dataset is derived from Matterport3D.It follows the Matterport End User License Agreement for Academic Use of Model Data.See Matterport3D License for details.
3DGS_Results3dgs-output3dgs-mipneft360-images23DGS_Tamper3dgs-dissolve-videosContains 101 videos of dissolving 3DGS (3D Gaussian Splatting) objects.
Splats are produced by running TRELLIS on 3D Arena, then dissolving the splats with VFX Graph in Unity.
hypersim_mcmc_3dgs
Data Statistics
Scenes
Mean PSNR ↑
Mean SSIM ↑
Mean LPIPS ↓
Mean Depth L1 ↓
Mean #3DGS
Total #3DGS
455
32.51 dB
0.942
0.078
0.0025 m
2.494M
1.135B
License Notice:This dataset is derived from the Hypersim source dataset and follows CC BY-SA 3.0.
bigym-all-tasks-3dgs-one-success
BiGym 全任务 3DGS 厨房壳 单成功轨迹集
发布状态：40/40 技术验证通过，视觉抽检通过；仓库保持，等待上游数据/壳资产再分发条款复核。
预览视频：previews/collection-preview.mp4
核心结果
官方 BiGym 任务：40/40
唯一任务：40
保存的 reward=1 episode：40
丢弃的 reward=0 候选：61（未进入成功 Parquet）
Transition / 每相机帧数：14,806 / 14,806
三相机 H.264 文件：120
相机：head 848×480、left wrist 640×480、right wrist 640×480，20 FPS
数据布局
reach/：3 个任务
long_horizon/：3 个任务
dishwasher/：10 个任务
tabletop/：24 个任务
task-manifest.csv：任务、demo UUID、seed、帧数、reward、动作哈希… See the full description on the dataset page: https://huggingface.co/datasets/eustance/bigym-all-tasks-3dgs-one-success.replica_mcmc_3dgs
Data Statistics
Scenes
Mean PSNR ↑
Mean SSIM ↑
Mean LPIPS ↓
Mean Depth L1 ↓
Mean #3DGS
Total #3DGS
18
39.43 dB
0.978
0.056
0.0027 m
1.500M
27.000M
License Notice:This dataset is derived from the Replica Dataset.It follows the Replica Dataset Research Terms of Use.See Replica License for details.
autoimagine-3dgs3dgs-dissolveCaptioned version of dylanebert/3dgs-dissolve-videos.
Captioning script:
caption.py
from transformers import Qwen2_5_VLForConditionalGeneration, AutoProcessor
import torch
import os
from pathlib import Path
from huggingface_hub import snapshot_download
from torchvision import io
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
"Qwen/Qwen2.5-VL-7B-Instruct",
device_map="auto",
torch_dtype=torch.bfloat16,
attn_implementation="flash_attention_2",
)
processor =… See the full description on the dataset page: https://huggingface.co/datasets/finetrainers/3dgs-dissolve.recondreamer-rl-stage2-3dgs-render-cache
ReConDreamer RL Stage2 3DGS Render Cache
Packed shards for render_cache_hysteresis002_all_fixed.
Restore on target machine:
mkdir -p /data/recondreamer-rl-stage2
cd /data/recondreamer-rl-stage2
for f in render_cache_hysteresis002_all_fixed_parts/*.tar.zst; do
tar --use-compress-program=unzstd -xf "$f"
done
Train with:
python policy/RAD/tools/train_rad_stage2_3dgs.py \
--dataset-root /data/recondreamer-rl-stage2/render_cache_hysteresis002_all_fixed
seu-3dgs
Single-Event Upsets in 3D Gaussian Splatting Rendering
Artifacts for the paper Single-Event Upsets in 3D Gaussian Splatting Rendering: Bit-Level Criticality, Spatial Extent, and a Parallel Support Guard (F. Alpay and B. Basaran).
A trained 3DGS model is a large floating-point array resident in GPU memory, so a single-event upset is one flipped bit in one parameter. This repository releases the fault-injection engine, the trained models, the per-cell aggregated records of more… See the full description on the dataset page: https://huggingface.co/datasets/Lightcap/seu-3dgs.3dgs3dgs_cldm_dataset_and_weight