WorkFlow
Shopping online for clothes and accessories is made more personal, by letting the Shoppers/Users try on the clothes visually. before actually ordering them. We use Augmented Reality technology on a users mobile browser so that , they don't actually have to download another app. Our product can work as a webpage. The user can click only one single image for personalising his mannequin to their size. Hence they can now easily try on the clothes with different color and size to see which fits the best and which design matches their personality well
we are using SMPLX[https://github.com/vchoutas/smplify-x] and OpenPose[https://github.com/CMU-Perceptual-Computing-Lab/openpose] for generating a custom 3D Human model for each users (virtual mannequin) from a single image(monocular) of a mobile camera.
To fit SMPL or SMPL+H, replace the yaml configuration file with either fit_smpl.yaml or fit_smplx.yaml, i.e.:
- for SMPL:
python smplifyx/main.py --config cfg_files/fit_smpl.yaml
--data_folder DATA_FOLDER
--output_folder OUTPUT_FOLDER
--visualize="True/False"
--model_folder MODEL_FOLDER
Follow the necessary steps to setup in given dependencies.
Visualizing the result from ZapparWorks[https://zap.works] {WebAR using three.js} where the 3d model is augmented in the real world with feature available [World Tracking/ Image trigger]
Follow the installation instructions for each of the following before using the fitting code.