超强图像分割SAM3 复现完整流程
本内容在博客同步发布喵七斗的猫猫海洋 0）环境前置
-
NVIDIA 独显
-
建议驱动最新（对应 CUDA 12.6+）
-
已装 Conda
-
注意：SAM3 与新版 opencv/numpy 可能冲突
我使用的是镜像 PyTorch 2.8.0 Python 3.12(ubuntu22.04) CUDA12.8 GPU：RTX 4090(24GB)*1
1）创建并激活环境
建议 Python 3.12
conda create -n sam3 python=3.12 -y conda activate sam3
2）安装 PyTorch（cu126）/AotuDL已有镜像选择则跳过
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu126
3）下载源码并安装
下载源码
git clone https://github.com/facebookresearch/sam3.git
进入解压后的目录：
cd path osam3-main pip install -e .
4）补齐依赖（未知原因导致包缺失）
pip install matplotlib pandas tqdm pillow pip install scikit-image scikit-learn
最终pip list 内容
Package Version Editable project location ------------------------- ----------- ------------------------- anyio 4.12.1 argon2-cffi 25.1.0 argon2-cffi-bindings 25.1.0 asttokens 3.0.0 async-lru 2.0.5 attrs 25.4.0 babel 2.17.0 beautifulsoup4 4.14.2 bleach 6.3.0 brotlicffi 1.2.0.0 certifi 2026.1.4 cffi 2.0.0 charset-normalizer 3.4.4 click 8.3.1 comm 0.2.3 contourpy 1.3.3 cuda-bindings 12.9.4 cuda-pathfinder 1.3.3 cycler 0.12.1 debugpy 1.8.16 decorator 5.2.1 decord 0.6.0 defusedxml 0.7.1 einops 0.8.1 executing 2.2.1 fastjsonschema 2.21.2 filelock 3.20.3 fonttools 4.61.1 fsspec 2026.1.0 ftfy 6.1.1 h11 0.16.0 hf-xet 1.2.0 html5lib 1.1 httpcore 1.0.9 httpx 0.28.1 huggingface_hub 1.3.3 idna 3.11 ImageIO 2.37.2 iopath 0.1.10 ipykernel 6.31.0 ipython 9.7.0 ipython_pygments_lexers 1.1.1 ipywidgets 8.1.7 jedi 0.19.2 Jinja2 3.1.6 joblib 1.5.3 json5 0.12.1 jsonschema 4.25.1 jsonschema-specifications 2025.9.1 jupyter 1.1.1 jupyter_client 8.8.0 jupyter-console 6.6.3 jupyter_core 5.9.1 jupyter-events 0.12.0 jupyter-lsp 2.2.5 jupyter_server 2.17.0 jupyter_server_terminals 0.5.3 jupyterlab 4.5.0 jupyterlab_pygments 0.3.0 jupyterlab_server 2.28.0 jupyterlab_widgets 3.0.16 kiwisolver 1.4.9 lazy_loader 0.4 MarkupSafe 3.0.3 matplotlib 3.10.8 matplotlib-inline 0.2.1 mistune 3.1.2 mpmath 1.3.0 nbclient 0.10.2 nbconvert 7.16.6 nbformat 5.10.4 nest-asyncio 1.6.0 networkx 3.6.1 notebook 7.5.0 notebook_shim 0.2.4 numpy 1.26.4 nvidia-cublas-cu12 12.8.4.1 nvidia-cuda-cupti-cu12 12.8.90 nvidia-cuda-nvrtc-cu12 12.8.93 nvidia-cuda-runtime-cu12 12.8.90 nvidia-cudnn-cu12 9.10.2.21 nvidia-cufft-cu12 11.3.3.83 nvidia-cufile-cu12 1.13.1.3 nvidia-curand-cu12 10.3.9.90 nvidia-cusolver-cu12 11.7.3.90 nvidia-cusparse-cu12 12.5.8.93 nvidia-cusparselt-cu12 0.7.1 nvidia-nccl-cu12 2.27.5 nvidia-nvjitlink-cu12 12.8.93 nvidia-nvshmem-cu12 3.4.5 nvidia-nvtx-cu12 12.8.90 opencv-python 4.11.0.86 packaging 26.0 pandas 3.0.0 pandocfilters 1.5.1 parso 0.8.5 pexpect 4.9.0 pillow 12.1.0 pip 25.3 platformdirs 4.5.0 portalocker 3.2.0 prometheus_client 0.21.1 prompt_toolkit 3.0.52 psutil 7.0.0 ptyprocess 0.7.0 pure_eval 0.2.3 pycocotools 2.0.11 pycparser 2.23 Pygments 2.19.2 pyparsing 3.3.2 PySocks 1.7.1 python-dateutil 2.9.0.post0 python-json-logger 4.0.0 PyYAML 6.0.3 pyzmq 27.1.0 qtconsole 5.7.0 QtPy 2.4.3 referencing 0.37.0 regex 2026.1.15 requests 2.32.5 rfc3339-validator 0.1.4 rfc3986-validator 0.1.1 rpds-py 0.28.0 safetensors 0.7.0 sam3 0.1.0 /root/autodl-tmp/sam3 scikit-image 0.26.0 scikit-learn 1.8.0 scipy 1.17.0 Send2Trash 1.8.3 setuptools 80.9.0 shellingham 1.5.4 six 1.17.0 sniffio 1.3.1 soupsieve 2.5 stack_data 0.6.3 sympy 1.14.0 terminado 0.18.1 threadpoolctl 3.6.0 tifffile 2026.1.14 timm 1.0.24 tinycss2 1.4.0 torch 2.10.0 torchvision 0.25.0 tornado 6.5.4 tqdm 4.67.1 traitlets 5.14.3 triton 3.6.0 typer-slim 0.21.1 typing_extensions 4.15.0 urllib3 2.6.3 wcwidth 0.3.5 webencodings 0.5.1 websocket-client 1.8.0 wheel 0.45.1 widgetsnbextension 4.0.14
5）权重下载
官方 HF 权限经常申请不到，几分钟给你拒了到也不占用时间（
sam3.pt链接：../images/2026/0206/tlib.pyplot as pltfrom PIL import Imagefrom sam3.model_builder import build_sam3_image_modelfrom sam3.model.sam3_image_processor import Sam3Processorfrom sam3.visualization_utils import plot_results").convert("RGB") # 全图编码 state = processor.set_image(image) # 文本提示分割（示例） state = processor.set_text_prompt(state=state, prompt="person") # 可视化 plot_results(image, state) # 1) 弹窗显示 # plt.show() # 2) 或者保存（可选） out_path = "out.png" plt.savefig(out_path, dpi=200, bbox_inches="tight", pad_inches=0) plt.close() print("Saved to:", os.path.abspath(out_path))
运行：
python main.py。
写在最后：
SAM3实例分割能力真的很强。（点头）而且4090跑起来很快（点头）
本文地址：https://www.yitenyun.com/6192.html
