Hugging Face's logo

schirrmacher
/
ormbg

ONNX
segmentation
remove background
background
background-removal
Pytorch
Model card Files Community
9
ormbg / README.md
schirrmacher's picture
schirrmacher
Upload ./README.md with huggingface_hub
63b502d verified
preview code
|
raw
history blame
3.33 kB
metadata
title: Open Remove Background Model (ormbg)
license: apache-2.0
tags:
  - segmentation
  - remove background
  - background
  - background-removal
  - Pytorch
pretty_name: Open Remove Background Model
models:
  - schirrmacher/ormbg
datasets:
  - schirrmacher/humans
emoji: 💻
colorFrom: red
colorTo: red
sdk: gradio
sdk_version: 4.29.0
app_file: hf_space/app.py
pinned: false

Open Remove Background Model (ormbg)

>>> DEMO <<<

Join our Research Discord Group!

This model is a fully open-source background remover optimized for images with humans. It is based on Highly Accurate Dichotomous Image Segmentation research. The model was trained with the synthetic Human Segmentation Dataset, P3M-10k, PPM-100 and AIM-500.

This model is similar to RMBG-1.4, but with open training data/process and commercially free to use.

Inference 

python ormbg/inference.py

Training

Install dependencies:

conda env create -f environment.yaml
conda activate ormbg

Replace dummy dataset with training dataset.

python3 ormbg/train_model.py

Research

I started training the model with synthetic images of the Human Segmentation Dataset crafted with LayerDiffuse. However, I noticed that the model struggles to perform well on real images.

Synthetic datasets have limitations for achieving great segmentation results. This is because artificial lighting, occlusion, scale or backgrounds create a gap between synthetic and real images. A "model trained solely on synthetic data generated with naïve domain randomization struggles to generalize on the real domain", see PEOPLESANSPEOPLE: A Synthetic Data Generator for Human-Centric Computer Vision (2022).

Next steps:

  • Expand dataset with synthetic and real images
  • Research on state of the art loss functions

Latest changes (26/07/2024):

  • Created synthetic dataset with 10k images, crafted with BlenderProc
  • Removed training data created with LayerDiffuse, since it lacks the accuracy needed
  • Improved model performance (after 100k iterations):

05/07/2024

  • Added P3M-10K dataset for training and validation
  • Added AIM-500 dataset for training and validation
  • Added PPM-100 dataset for training and validation
  • Applied Grid Dropout to make the model smarter