license: apache-2.0
pipeline_tag: image-segmentation
tags:
- model_hub_mixin
- pytorch_model_hub_mixin
EdgeCrafter: Compact ViTs for Edge Dense Prediction
EdgeCrafter is a unified framework for compact Vision Transformers (ViTs) designed for high-performance dense prediction (detection, instance segmentation, and pose estimation) on resource-constrained edge devices. This specific model, ECSeg-S, is a lightweight instance segmentation model.
- Paper: EdgeCrafter: Compact ViTs for Edge Dense Prediction via Task-Specialized Distillation
- GitHub Repository: Intellindust-AI-Lab/EdgeCrafter
- Project Page: EdgeCrafter Project Page
Model Description
ECSeg-S is built using a distilled compact backbone and an edge-friendly encoder-decoder design. It achieves a strong accuracy-efficiency tradeoff, making it suitable for real-time applications on edge hardware. For instance segmentation, it achieves performance comparable to RF-DETR while using significantly fewer parameters.
Quick Start (Inference)
To run inference on a sample image, follow the instructions from the official repository:
1. Installation
# Create conda environment
conda create -n ec python=3.11 -y
conda activate ec
# Install dependencies
pip install -r requirements.txt
2. Run Inference
# Navigate to the detection/segmentation folder
cd ecdetseg
# Run PyTorch inference
# Replace `path/to/your/image.jpg` with an actual image path
python tools/inference/torch_inf.py -c configs/ecseg/ecseg_s.yml -r /path/to/ecseg_s.pth -i path/to/your/image.jpg
Citation
If you find EdgeCrafter useful in your research, please consider citing:
@article{liu2026edgecrafter,
title={EdgeCrafter: Compact ViTs for Edge Dense Prediction via Task-Specialized Distillation},
author={Liu, Longfei and Hou, Yongjie and Li, Yang and Wang, Qirui and Sha, Youyang and Yu, Yongjun and Wang, Yinzhi and Ru, Peizhe and Yu, Xuanlong and Shen, Xi},
journal={arXiv},
year={2026}
}