3D_generator

This is a locally deployed web application for generating 3D models from single images. It integrates InstructPix2Pix for image fine-tuning using text prompts and supports 3D model generation in .obj format. In addition, a database is implemented to enable similar model retrieval through CLIP feature search.

Installation

Create environment by

conda env create -f environment.yml

Update setuptools by

pip install --upgrade setuptools

Install other dependencies by

pip install -r requirements.txt

Download a pretrained model for instructionPix2Pix

bash scripts/download_checkpoints.sh

Manual install dlib

git clone https://github.com/Murtaza-Saeed/Dlib-Precompiled-Wheels-for-Python-on-Windows-x64-Easy-Installation.git
cd Dlib-Precompiled-Wheels-for-Python-on-Windows-x64-Easy-Installation

Step 1: Install CMake Run the following command in your terminal or command prompt:

pip install cmake

Step 2: Install Dlib Download the appropriate .whl file for your Python version from this repository. Navigate to the folder where the file is located using the command prompt, and run:

pip install <filename>

Replace <filename> with the name of the downloaded .whl file.

Example Commands

• For Python 3.7:

  python -m pip install dlib-19.22.99-cp37-cp37m-win_amd64.whl

• For Python 3.8:

  python -m pip install dlib-19.22.99-cp38-cp38-win_amd64.whl

• For Python 3.9:

  python -m pip install dlib-19.22.99-cp39-cp39-win_amd64.whl

• For Python 3.10:

  python -m pip install dlib-19.22.99-cp310-cp310-win_amd64.whl

• For Python 3.11:

  python -m pip install dlib-19.24.1-cp311-cp311-win_amd64.whl

• For Python 3.12:

  python -m pip install dlib-19.24.99-cp312-cp312-win_amd64.whl

Manual install k-diffusion

pip install git+https://github.com/crowsonkb/k-diffusion.git

Run the demo website

python ./demo_app.py