1

.gitignore

@@ -0,0 +1,5 @@
+*.pth
+Qwen2-VL-2B-Instruct
+.venv
+.idea
+system.json

Dockerfile

@@ -0,0 +1,43 @@
+FROM ghcr.io/astral-sh/uv:latest AS uv_bin
+FROM nvidia/cuda:12.4.1-base-ubuntu22.04
+
+ENV UV_LINK_MODE=copy \
+    UV_COMPILE_BYTECODE=1 \
+    PYTHONUNBUFFERED=1 \
+    UV_PROJECT_ENVIRONMENT=/app/.venv \
+    TORCH_CUDA_ARCH_LIST="8.6" \
+    CUDA_VISIBLE_DEVICES=0 \
+    UV_TORCH_BACKEND=cu128
+
+COPY --from=uv_bin /uv /uvx /bin/
+
+#RUN apt-get update && apt-get install -y --no-install-recommends \
+#    ca-certificates curl gnupg && \
+#    sed -i 's|http://archive.ubuntu.com/ubuntu/|http://mirrors.cloud.tencent.com/ubuntu/|g' /etc/apt/sources.list && \
+#    sed -i 's|http://security.ubuntu.com/ubuntu/|http://mirrors.cloud.tencent.com/ubuntu/|g' /etc/apt/sources.list && \
+#    apt-get update && \
+#    apt-get install -y --no-install-recommends \
+#    git \
+#    ffmpeg libsm6 libxext6 \
+#    build-essential g++ \
+#    && apt-get clean && rm -rf /var/lib/apt/lists/*
+
+WORKDIR /app
+
+# 3. 安装依赖 (不加 --system，让 uv 创建受管的虚拟环境)
+# 这里会根据 pyproject.toml 自动下载并安装 Python 3.11
+COPY pyproject.toml uv.lock ./
+#RUN uv sync --frozen --no-dev --no-install-project --python 3.9
+# 4. 拷贝项目文件并安装项目本身
+COPY . .
+#RUN uv sync --frozen --no-dev --python 3.9
+
+# 5. 【最关键】将虚拟环境的 bin 目录提到最前面
+# 注意：uv sync 创建的 python 就在这个目录下
+ENV PATH="/app/.venv/bin:$PATH"
+
+EXPOSE 8000
+
+# 验证路径并运行
+# 此时运行 python 实际上是运行 /app/.venv/bin/python
+CMD ["sleep", "infinity"]

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 Style3D
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -1,23 +1,29 @@
 
-# Design2GarmentCode:  Turning Design Concepts to Tangible Garments Through Program Synthesis
+# Design2GarmentCode: Turning Design Concepts to Tangible Garments Through Program Synthesis
 
+[![arXiv](https://img.shields.io/badge/📃-arXiv%20-red.svg)](https://arxiv.org/abs/2412.08603)
+[![webpage](https://img.shields.io/badge/🌐-Website%20-blue.svg)](https://style3d.github.io/design2garmentcode/) 
+[![Youtube](https://img.shields.io/badge/📽️-Video%20-orchid.svg)](https://www.youtube.com/xxx)
 
-[arXiv](https://arxiv.org/abs/2412.08603) | [Project Page](https://style3d.github.io/design2garmentcode/)
+<span class="author-block"><a href="">Feng Zhou</a>,&nbsp;</span>
+<span class="author-block"><a href="https://walnut-ree.github.io/">Ruiyang Liu</a>,&nbsp;</span>
+<span class="author-block"><a href="">Chen Liu</a>,&nbsp;</span>
+<span class="author-block"><a href="">Gaofeng He</a>,&nbsp;</span>
+<span class="author-block"><a href="https://dirtyharrylyl.github.io/">Yong-Lu Li</a>,&nbsp;</span>
+<span class="author-block"><a href="http://www.cad.zju.edu.cn/home/jin/">Xiaogang Jin</a>,&nbsp;</span>
+<span class="author-block"><a href="https://wanghmin.github.io/">Huamin Wang</a></span>
 
-Feng Zhou, Ruiyang Liu, Chen Liu, Gaofeng He, Yong‑Lu Li, Xiaogang Jin, Huamin Wang. *CVPR 2025 .*
+<p align="center">
+  <img src="https://github.com/Style3D/design2garmentcode-impl/raw/main/assets/img/neural_symbolic-pipeline.png">
+</p>
+Official implementation for Design2GarmentCode, a motility-agnostic sewing pattern generation framework that leverages fine-tuned Large Multimodal Models to generate parametric pattern-making programs from multi-modal design concepts.
 
-![teaser](assets/img/neural_symbolic-pipeline.png)
- we propose a novel
-sewing pattern generation approach Design2GarmentCode
-based on Large Multimodal Models (LMMs), to generate parametric pattern-making programs from multi-modal
-design concepts
----
 
 ## Installation
 ### 1. Clone the repository
 ```bash
-git clone https://github.com/your-org/design2garmentcode.git   # ← replace with the real URL
-cd design2garmentcode
+git clone https://github.com/Style3D/design2garmentcode-impl.git
+cd design2garmentcode-impl
 ```
 
 ### 2. Create the Conda environment
@@ -38,57 +44,30 @@ If you need local cloth simulation and 3‑D visualization, follow the installat
 `Design2GarmentCode` communicates with large multimodal models.  
 Follow the steps **in the given order**:
 
-1. **Provide API credentials**  
-   - **Environment variable (recommended)** – defaults to *ChatGPT‑4o*  
+#### 1. **Provide API credentials for MMUA**  
+- **Environment variable (recommended)** – defaults to *ChatGPT‑4o*  
      ```bash
      export OPENAI_API_KEY="sk‑..."
      ```  
-   - **Edit `system.json`** (project root) – manually specify `api_key`, `base_url`, and `model` if you prefer a file‑based approach.
-2. **Download the required models**:  
-   - First, download the base model [Qwen2-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct/tree/main).  
-     Place the entire folder at:  
-     `lmm_utils/Qwen/Qwen2-VL-2B-Instruct/`
+- **Edit `system.json`** (project root) – manually specify `api_key`, `base_url`, and `model` if you prefer a file‑based approach.
 
-   - Next, download the fine-tuned weights file [model.pth](lmm_utils/Qwen/qwen2vl_lora_mlp/model.pth),  
-     and place it in:  
-     `lmm_utils/Qwen/qwen2vl_lora_mlp/`
+#### 2. **Set up the parameter projector**:  
+- Download the base model [Qwen2-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct/tree/main) and place the modal to `lmm_utils/Qwen/Qwen2-VL-2B-Instruct/`.
+
+- Download the fine-tuned weights file from [Google Drive](https://drive.google.com/file/d/1CL7OLUq6fYcwoDuLRkBxtKNxJ0_G73U-/view?usp=sharing), and place it in `lmm_utils/Qwen/qwen2vl_lora_mlp/`.
 ---
 
-## Quick GUI Demo
+## Testing with GUI
 
-```bash
-python gui.py         
-```
-![GUI Demo](assets/img/gui.png)
-- Input: Free‑form prompt or an image/sketch  
-- Output: GarmentCode JSON, preview image, and (optionally) physics simulation
+Setting up the GUI with `python gui.py` where you will see the following interface (modified from GarmentCode)
 
----
+<p align="center">
+  <img src="https://github.com/Style3D/design2garmentcode-impl/raw/main/assets/img/gui_example.png">
+</p>
 
-### 1. Text-Guided Pattern Generation
+Switching to the `PARSE DESIGN` tab, and input your design input, either text description, photograph or sketch, to the chatbox. The generated sewing pattern will appear on the right side after parsing.
 
-- Go to the PARSE DESIGN tab.
-- In the input box at the bottom ("Describe your design..."), type a natural language description of the garment.  
-  e.g., a T-shirt
-- Click SEND to generate patterns based on your description.
-
----
-
-### 2. Image-Guided Pattern Generation
-
-- Click the upload icon inside the input box to upload a reference image or sketch.
-- Once the image is uploaded, click SEND to parse the design and generate corresponding patterns.
-
----
-
-### 3. Modify Patterns in the GUI
-
-Once a pattern is generated, you can refine it directly inside the GUI:
-
-1. Focus the input box at the bottom.  
-2. Type `modify: <your-instruction>`  
-   - e.g., `modify: make sleeves shorter`  
-3. Press Enter – the system will regenerate the pattern accordingly.
+Once a pattern is generated, you can modify the result by typing `modify: <your-instruction>` in the chatbox.
 
 ---
 ## Batch Inference
@@ -126,27 +105,23 @@ python lmm_utils/test_picture_batch.py \
 
 ---
 
-
-## Get 3D Garment Patterns
+## Simulate 3D Garment
 ### 1. Generate from a pattern.json
-After generating the pattern data, you can simulate the corresponding 3D output directly from the pattern's JSON file.
+After generating the pattern data, you can simulate the corresponding 3D output directly from the pattern's JSON file with
 ```bash
 python test_garment_sim.py --pattern_spec $INPUT_JSON 
 ```
-### 2. Generate from gui
-You can also run the simulation directly on the GUI to obtain 3D data.
-```bash
-python gui.py 
-```
+Or run the simulation directly in the `3D View` GUI tab.
+
 ### Citation
-```bash
 If you find this work useful, please cite:
 
 ```bibtex
-@article{zhou2024design2garmentcode,
+@inproceedings{zhou2025design2garmentcode,
   title={Design2GarmentCode: Turning Design Concepts to Tangible Garments Through Program Synthesis},
   author={Zhou, Feng and Liu, Ruiyang and Liu, Chen and He, Gaofeng and Li, Yong-Lu and Jin, Xiaogang and Wang, Huamin},
-  journal={arXiv preprint arXiv:2412.08603},
-  year={2024}
+  booktitle={Proceedings of the Computer Vision and Pattern Recognition Conference},
+  pages={23712--23722},
+  year={2025}
 }
 ```

docker-compose.yaml

@@ -0,0 +1,25 @@
+services:
+  server:
+    container_name: aida_design2garment
+    build:
+      context: .
+      dockerfile: Dockerfile
+    working_dir: /app
+    volumes:
+      - .:/app
+    ports:
+      - "10080:8000"
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              device_ids: [ '1' ]
+              capabilities:
+                - gpu
+    networks:
+      - aida_app_net
+networks:
+  aida_app_net:
+    external: true
+    name: aida_app_net

pyproject.toml

@@ -0,0 +1,32 @@
+[project]
+name = "design2garmentcode-impl"
+version = "0.1.0"
+description = "Add your description here"
+readme = "README.md"
+requires-python = ">=3.9"
+dependencies = [
+    "accelerate==1.1.1",
+    "cairosvg==2.7.1",
+    "cgal==6.0.1.post202410241521",
+    "datasets==2.18.0",
+    "huggingface-hub==0.29.2",
+    "libigl==2.5.1",
+    "markupsafe==2.1.5",
+    "matplotlib>=3.9.4",
+    "modelscope==1.18.0",
+    "nicegui==2.15.0",
+    "numpy==1.26.4",
+    "peft==0.13.2",
+    "psutil==6.0.0",
+    "pyrender==0.1.45",
+    "pyyaml==6.0.2",
+    "qwen-vl-utils==0.0.8",
+    "safetensors==0.5.3",
+    "scipy==1.13.1",
+    "svgpathtools>=1.7.2",
+    "svgwrite==1.4.3",
+    "tiktoken==0.9.0",
+    "tokenizers==0.20.3",
+    "transformers==4.46.2",
+    "trimesh>=4.11.5",
+]

requirements.txt

@@ -0,0 +1,29 @@
+- --extra-index-url https://download.pytorch.org/whl/cu121
+- numpy==1.26.4
+- scipy==1.13.1
+- pyyaml==6.0.2
+- svgwrite==1.4.3
+- psutil==6.0.0
+- matplotlib
+- svgpathtools
+- cairosvg==2.7.1
+- nicegui==2.15.0
+- trimesh
+- cgal
+- torch==2.4.0+cu121
+- torchvision==0.19.0+cu121
+- torchaudio==2.4.0+cu121
+- transformers==4.46.2
+- tokenizers==0.20.3
+- accelerate==1.1.1
+- datasets==2.18.0
+- huggingface-hub==0.29.2
+- safetensors==0.5.3
+- tiktoken==0.9.0
+- peft==0.13.2
+- qwen-vl-utils==0.0.8
+- modelscope==1.18.0
+- pyrender==0.1.45
+- libigl==2.5.1
+- cgal==6.0.1.post202410241521
+- openai==1.54.4