Merge branch 'master' of github.com:cre185/InstantSfM

Author: cre185

.gitignore

@@ -186,4 +186,5 @@ debug_dir/*
 debug_output/*
 test
 tmp/*
-demo_tmp/
+demo_tmp/
+examples/*

Dockerfile

@@ -0,0 +1,74 @@
+FROM nvcr.io/nvidia/pytorch:25.11-py3
+
+ARG TARGETARCH
+ARG DEBIAN_FRONTEND=noninteractive
+
+# Core build tooling and runtime libs needed by scikit-sparse / OpenCV / fused-ssim
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    build-essential cmake ninja-build git wget ca-certificates xz-utils \
+    libgoogle-glog-dev libgflags-dev libatlas-base-dev libeigen3-dev \
+    libsuitesparse-dev libmetis-dev liblapack-dev libblas-dev \
+    libgl1 libglib2.0-0 colmap \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install cuDSS for CUDA 13 (arch-aware)
+ARG CUDSS_VERSION=0.7.1.4
+ENV CUDSS_ROOT=/opt/cudss-${CUDSS_VERSION}
+RUN set -euo pipefail; \
+    arch="${TARGETARCH:-$(uname -m)}"; \
+    case "${arch}" in \
+        amd64|x86_64) cudss_arch="x86_64"; deb_arch_dir="x86_64-linux-gnu";; \
+        arm64|aarch64) cudss_arch="aarch64"; deb_arch_dir="aarch64-linux-gnu";; \
+        *) echo "Unsupported architecture: ${arch}"; exit 1;; \
+    esac; \
+    CUDSS_TARBALL="libcudss-linux-${cudss_arch}-${CUDSS_VERSION}_cuda13-archive.tar.xz"; \
+    CUDSS_URL_BASE="https://developer.download.nvidia.com/compute/cudss/redist/libcudss/linux-${cudss_arch}"; \
+    wget -q "${CUDSS_URL_BASE}/${CUDSS_TARBALL}" -O /tmp/cudss.tar.xz; \
+    tar -xJf /tmp/cudss.tar.xz -C /opt; \
+    ln -s "/opt/libcudss-linux-${cudss_arch}-${CUDSS_VERSION}_cuda13-archive" "${CUDSS_ROOT}"; \
+    mkdir -p /usr/include/libcudss/12 "/usr/lib/${deb_arch_dir}/libcudss/12"; \
+    cp -r "${CUDSS_ROOT}/include/"* /usr/include/libcudss/12/; \
+    cp -r "${CUDSS_ROOT}/lib/"* "/usr/lib/${deb_arch_dir}/libcudss/12/"; \
+    ldconfig; \
+    rm /tmp/cudss.tar.xz
+
+# cuDSS toolchain paths
+ENV CUDSS_LIBCUDSS_PATHS=/usr/lib/aarch64-linux-gnu/libcudss/12:/usr/lib/x86_64-linux-gnu/libcudss/12
+ENV LD_LIBRARY_PATH=${CUDSS_ROOT}/lib:${CUDSS_LIBCUDSS_PATHS}:${LD_LIBRARY_PATH}
+ENV LIBRARY_PATH=${CUDSS_ROOT}/lib:${CUDSS_LIBCUDSS_PATHS}:${LIBRARY_PATH}
+ENV CPATH=${CUDSS_ROOT}/include
+ENV CUDA_HOME=/usr/local/cuda
+ENV FUSED_SSIM_FORCE_CUDA=1
+ENV TORCH_CUDA_ARCH_LIST="8.0;9.0;8.6;8.9;11.0"
+
+# Build and install Ceres Solver (needed by pyceres)
+ARG CERES_VERSION=2.1.0
+RUN wget -q http://ceres-solver.org/ceres-solver-${CERES_VERSION}.tar.gz \
+    && tar zxf ceres-solver-${CERES_VERSION}.tar.gz \
+    && mkdir ceres-build \
+    && cd ceres-build \
+    && cmake ../ceres-solver-${CERES_VERSION} -GNinja \
+         -DBUILD_TESTING=OFF -DBUILD_EXAMPLES=OFF -DBUILD_SHARED_LIBS=ON \
+         -DMINIGLOG=OFF -DSUITESPARSE=OFF -DCXSPARSE=OFF \
+    && ninja install \
+    && cd / \
+    && rm -rf ceres-build ceres-solver-${CERES_VERSION} ceres-solver-${CERES_VERSION}.tar.gz
+
+WORKDIR /workspace/InstantSfM
+
+# Install Python dependencies (including external git deps) first for better layer caching
+COPY pyproject.toml README.md ./
+COPY instantsfm instantsfm
+COPY demo.py demo.py
+RUN pip install --no-cache-dir numpy==1.26.4 \
+    && git clone --depth=1 https://github.com/rahul-goel/fused-ssim /tmp/fused-ssim \
+    && python -c "from pathlib import Path; p=Path('/tmp/fused-ssim/setup.py'); t=p.read_text(); t=t.replace('elif torch.mps.is_available():','elif False and torch.mps.is_available():'); t=t.replace('elif hasattr(torch, \\'xpu\\'):', 'elif False and hasattr(torch, \\'xpu\\'):'); p.write_text(t)" \
+    && pip install --no-cache-dir --no-build-isolation /tmp/fused-ssim \
+    && pip install --no-cache-dir --upgrade pip \
+    && MAX_JOBS=$(nproc) pip install --no-cache-dir --no-build-isolation git+https://github.com/zitongzhan/bae.git \
+    && pip install --no-cache-dir .
+
+# Bring in the rest of the project (assets, configs, etc.)
+COPY . .
+
+CMD ["/bin/bash"]

README.md

@@ -1,4 +1,4 @@
-<img src="img/USC-Logos.png" width=120px /><img src="img/University_at_Buffalo_logo.png" width=120px /><img src="img/Tsinghua_University_Logo.png" width=50px />
+<img src="https://github.com/cre185/InstantSfM/blob/gh-pages/static/images/USC-Logos.png?raw=true" width=120px /><img src="https://github.com/cre185/InstantSfM/blob/gh-pages/static/images/University_at_Buffalo_logo.png?raw=true" width=120px /><img src="https://github.com/cre185/InstantSfM/blob/gh-pages/static/images/Tsinghua_University_Logo.png?raw=true" width=50px />
 
 <div align="center">
 
@@ -36,6 +36,7 @@
 **⚠️Please note that this repository is still under active development. We will keep updating it regularly. Feel free to open an issue if you encounter any problem.**
 
 ## News📰  
+- **2025/12/02**: Added a Dockerfile and quick test command to run the bundled `examples/kitchen` dataset.  
 - **2025/11/27**: We changed the data structure into a more SIMD-friendly format, which further speeds up the whole pipeline by around 10%.  
 
 ## 1. Installation  
@@ -102,6 +103,21 @@ To run the demo, simply try the command `python demo.py`. In the demo, you can c
 ```
 In both cases, the output will be saved in the corresponding folder(`demo_output/` or your specified folder), and the results will be displayed directly in the web viewer.  
 
+## Docker quick test (kitchen example)
+If you built the provided `Dockerfile` into an image tagged `instantsfm`, 
+you can sanity‑check the pipeline on the bundled `examples/kitchen` data with a single container run. This Dockerfile supports both arm64 and x86_64 architectures. Build the Docker image with:
+```bash
+docker build -t instantsfm .
+```
+The command below mounts the repo so results persist to your host and uses `--rm` for a clean exit:
+```bash
+docker run --rm --gpus all --ipc=host --ulimit memlock=-1 --ulimit stack=67108864 \
+  -v "$PWD":/workspace/InstantSfM -w /workspace/InstantSfM instantsfm \
+  bash -lc "ins-feat --data_path examples/kitchen --feature_handler colmap --manual_config_name colmap && \
+            ins-sfm --data_path examples/kitchen --manual_config_name colmap --export_txt"
+```
+Outputs will appear under `examples/kitchen/sparse` on the host.
+
 ## 3. Command Line Usage
 The whole pipeline consists of three main steps: feature extraction and matching, global structure from motion (SfM), and 3DGS training.  
 Before performing these steps, prepair your dataset (a collection of images) in a folder structure like mentioned in the demo section, that is, a folder containing a subfolder `images/` with all the images inside.