站内导读
定位:Idol 是 Sperid Labs 发布的开源图像到 3D 模型,输入单张物体图片即可生成标准 3DGS 点云,适合快速原型和 3D 内容创作。
核心能力
- 单图生成最多 1,048,576 个高斯点,输出标准 3DGS .ply 文件,可直接用常见查看器打开。
- 内置 BiRefNet 去背景与自动裁剪,按 synthetic/real 模式调整物体占画面比例。
- 在消费级 GPU 上峰值显存 7.2 GB,采样可低于 2 秒;默认启用 torch.compile 等加速,约 1.7 倍提速。
- 提供 --seed、--steps、--cfg_strength、--num_gaussians 等参数控制生成细节与随机性。
适用人群:需要从单张图片快速获得 3D 高斯点云的研究者与开发者;需 Python 3.10+、PyTorch 2.5+、NVIDIA CUDA GPU,并先申请两个 gated 模型权限。
注意:首次运行需下载约 3 GB 权重及两个图像编码器,未获权限会提前停止。
位置づけ:Idol は Sperid Labs が公開した画像から 3D へのモデルで、物体の写真 1 枚から標準 3DGS の点群を生成し、プロトタイピングや 3D コンテンツ制作に使えます。
主な機能
- 1 枚の画像から最大 1,048,576 個のガウシアンを生成し、標準 3DGS の .ply として出力します。
- BiRefNet による背景除去と自動クロップを内蔵し、synthetic/real モードで物体の占有率を切り替えます。
- コンシューマー向け GPU でピーク VRAM 7.2 GB、サンプリングは 2 秒未満も可能です。
- torch.compile などによる高速化が既定で有効になり、約 1.7 倍の速度向上が見込めます。
対象ユーザー:1 枚の画像から手早く 3D ガウシアン点群を得たい研究者・開発者。Python 3.10+、PyTorch 2.5+、NVIDIA CUDA GPU が必要で、2 つの gated モデルの利用申請が前提です。
注意:初回実行時に約 3 GB の重みと 2 つの画像エンコーダをダウンロードします。
정의: Idol은 Sperid Labs가 공개한 이미지-투-3D 모델로, 물체 사진 한 장에서 표준 3DGS 포인트 클라우드를 생성해 빠른 프로토타이핑과 3D 콘텐츠 제작에 쓸 수 있습니다.
핵심 기능
- 이미지 한 장으로 최대 1,048,576개의 가우시안을 생성하고 표준 3DGS .ply로 출력합니다.
- BiRefNet 기반 배경 제거와 자동 크롭을 내장하고 synthetic/real 모드로 물체 점유율을 조정합니다.
- 소비자용 GPU에서 피크 VRAM 7.2GB, 샘플링은 2초 미만도 가능합니다.
- torch.compile 등 가속이 기본 활성화되어 약 1.7배 빨라집니다.
대상 사용자: 이미지 한 장에서 빠르게 3D 가우시안 포인트 클라우드를 얻으려는 연구자와 개발자. Python 3.10+, PyTorch 2.5+, NVIDIA CUDA GPU가 필요하며 두 gated 모델 승인이 선행되어야 합니다.
주의: 첫 실행 시 약 3GB 가중치와 두 이미지 인코더를 내려받습니다.
What it is: Idol is an open image-to-3D model from Sperid Labs that turns a single object picture into a standard 3D Gaussian splat, useful for quick prototyping and 3D content work.
Key capabilities
- Generates up to 1,048,576 Gaussians from one image and writes a standard 3DGS .ply viewable in common splat viewers.
- Built-in BiRefNet background removal and automatic cropping, with synthetic/real modes that change how much of the frame the object fills.
- Runs on consumer GPUs at 7.2 GB peak VRAM, with sampling possible under 2 seconds.
- Faster sampling is on by default via torch.compile, batched guided/unguided passes and cuDNN attention, about 1.7x speedup.
Who it is for: Researchers and developers who need a 3D Gaussian splat from a single image; requires Python 3.10+, PyTorch 2.5+, an NVIDIA CUDA GPU, and access to two gated Hugging Face models.
Caveats: The first run downloads about 3 GB of weights plus two image encoders; it stops early if encoder access is missing.
README · 详细文档
来源 · 查看原文
license: apache-2.0 libraryname: idol pipelinetag: image-to-3d tags:
- gaussian-splatting
- 3d
- image-to-3d
Idol
A single image to a 3D Gaussian splat.
One picture of an object in, up to a million 3D Gaussians out, including the sides the picture never showed.
Idol turns one picture of an object into a 3D Gaussian splat. Give it a render, a generated image or a photo: it removes the background, frames the object and generates up to 1,048,576 Gaussians, written as a standard 3DGS .ply. It runs on consumer GPUs, using 7.2 GB of VRAM and can run in under 2 seconds.
Examples
Idol turns one picture of an object into a 3D Gaussian splat. Give it a render, a generated image or a photo: it removes the background, frames the object and generates up to 1,048,576 Gaussians, including the sides the picture never showed. It runs on consumer GPUs, using 7.2 GB of VRAM and can run in under 2 seconds.
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Get started
1. Get access to the image encoders. Idol reads its input through two gated models on Hugging Face. Accept their terms with your Hugging Face account, then log in once with hf auth login (or set HF_TOKEN):
- facebook/dinov3-vith16plus-pretrain-lvd1689m
- black-forest-labs/FLUX.2-dev (only its VAE is downloaded)
2. Install the code
git clone https://github.com/speridlabs/idol-sdk.git
cd idol-sdk
pip install -e . # Python 3.10+, PyTorch 2.5+, an NVIDIA GPU with CUDA
There is no separate download step: on the first run, Idol's weights (about 3 GB, from this repository) and the two image encoders download automatically into the Hugging Face cache. If the encoders are not accessible yet, Idol stops before downloading anything and says what is missing.
3. Generate a 3D object
python -m idol.sample --image photo.jpg --mode real
python -m idol.sample --image renders/ --mode synthetic --out outputs
For each image this writes <name>.ply (open it in any Gaussian splat viewer), <name>_input.png (the image the model saw, after background removal and cropping) and <name>_camera.json (the predicted camera of the input).
Pick the mode for your input. It sets how much room the crop leaves around the object:
--mode | for | crop |
|---|---|---|
synthetic (default) | renders and generated images of a single object | object fills 1/1.2 of the frame |
real | photographs | object fills 1/1.7 of the frame |
From Python:
from idol import IdolPipeline
pipeline = IdolPipeline.from_pretrained("speridlabs/idol")
output = pipeline("photo.jpg", mode="real", seed=42, num_gaussians=1_048_576)
output.gaussian.save_ply("photo.ply")
Flags
| Flag | Default | What it does |
|---|---|---|
--seed | 42 | Fix it to get the same object again. |
--steps | 50 | Sampling steps. Fewer are faster but lose some detail. |
--cfg_strength | 3.0 | Classifier-free guidance: how strictly the object follows the image. |
--rescale_t | 3.0 | Time shift: higher spends more steps on the overall shape, lower on fine detail. |
--num_gaussians | 262144 | 32,768 to 1,048,576, a multiple of 32. |
--dinov3_dtype | bf16 | fp32 runs the DINOv3 encoder in full precision (1.7 GB more memory). |
--weights | speridlabs/idol | A Hugging Face repo or a local folder. |
--no_optimize | optimized | Turn off the faster sampling. |
Faster sampling is on by default on GPU (~1.7x, same output quality): the denoiser is compiled with torch.compile, the guided and unguided passes run as one batch, and attention uses cuDNN with TF32 matmuls while sampling. The first image in a process is slower while it compiles.
More details about the model
| Component | Parameters | Precision on GPU | Weights in VRAM |
|---|---|---|---|
| DiT (denoiser) | 370.6M | FP16 transformer blocks; FP32 embeddings and norms | ~0.8–1.0 GB |
| Octree decoder | 50.8M | FP16 transformer blocks; FP32 embeddings and norms | ~0.1–0.2 GB |
| Gaussian decoder | 237.3M | FP16 transformer blocks; FP32 embeddings and norms | ~0.5–0.9 GB |
| BiRefNet (background removal) | 220.7M | FP16 | ~0.44 GB |
| DINOv3 ViT-H+/16 | 840.6M | bf16 | ~1.68 GB |
| FLUX.2 VAE | 84.1M | bf16 (FP32 checkpoint, cast on load) | ~0.17 GB |
| Full request | ~1.80B | mostly FP16 / bf16, some FP32 | ~3.7–4.4 GB of weights, 7.2 GB peak |
Benchmark
Idol, TripoSplat, TRELLIS and LGM, scored at the input view (PSNR, SSIM, LPIPS, IoU) and at seven new views (DINOv2, CLIP, sharpness distance). Mean scores, 1 × H100 per model. Bold is the best in each column.
| Model | Setting | PSNR ↑ | SSIM ↑ | LPIPS ↓ | IoU ↑ | DINOv2 ↑ | CLIP ↑ | Sharpness distance ↓ | Request time ↓ |
|---|---|---|---|---|---|---|---|---|---|
| Idol | 50 steps, 262,144 Gaussians | 12.25 | 0.726 | 0.213 | 0.862 | 0.781 | 0.885 | 0.81 | 6.8 s |
| TripoSplat | 50 steps, 262,144 Gaussians | 11.76 | 0.712 | 0.225 | 0.860 | 0.743 | 0.868 | 1.03 | 9.9 s |
| TRELLIS | Paper setting, 50 + 50 steps | 11.89 | 0.716 | 0.235 | 0.845 | 0.754 | 0.868 | 0.97 | 7.1 s |
| LGM | Default, 65,536 Gaussians | 12.46\* | 0.722 | 0.237 | 0.834 | 0.575 | 0.784 | 1.22 | 1.1 s |
\* LGM has the highest PSNR despite visibly blurry results, because PSNR averages squared pixel error.
Across GPUs
One image to 262,144 Gaussians in 3.4 s. Request time at 50 steps: median over 8 images, with the same code and settings on every GPU, one GPU per machine.
| GPU | VRAM | Request time |
|---|---|---|
| B300 | 269 GB | 3.4 s |
| B200 | 179 GB | 3.6 s |
| H200 | 140 GB | 6.6 s |
| H100 | 80 GB | 6.7 s |
| RTX 5090 | 32 GB | 14.1 s |
| A100 | 80 GB | 16.3 s |
| L40S | 45 GB | 16.6 s |
| RTX 4090 | 24 GB | 19.1 s |
| RTX 3080 | 10 GB | 46.2 s |
| L4 | 22 GB | 54.6 s |
Made with ❤️ by speridlabs.com







