Skip to content

SAHI: Slicing Aided Hyper Inference

A lightweight vision library for performing large scale object detection & instance segmentation

teaser

downloads downloads License pypi version conda version Continious Integration
Context7 MCP llms.txt ci arXiv Open In Colab HuggingFace Spaces

What is SAHI?

SAHI (Slicing Aided Hyper Inference) is an open-source framework that provides a generic slicing-aided inference and fine-tuning pipeline for small object detection. Detecting small objects and those far from the camera is a major challenge in surveillance applications, as they are represented by a small number of pixels and lack sufficient detail for conventional detectors.

SAHI addresses this by applying a unique methodology that can be used with any object detector without requiring additional fine-tuning. Experimental evaluations on the Visdrone and xView aerial object detection datasets show that SAHI can increase object detection AP by up to 6.8% for FCOS, 5.1% for VFNet, and 5.3% for TOOD detectors. With slicing-aided fine-tuning, the accuracy can be further improved, resulting in a cumulative increase of 12.7%, 13.4%, and 14.5% AP, respectively. The technique has been successfully integrated with Detectron2, MMDetection, and YOLOv5 models.

  •   Getting Started

    Install sahi with pip and get up and running in minutes.

    Quickstart

  •   Predict

    Predict on new images, videos and streams with SAHI.

    Learn more

  •   Slicing

    Learn how to slice large images and datasets for inference.

    Learn more

  •   COCO Utilities

    Work with COCO format datasets, including creation, splitting, and filtering.

    Learn more

  •   CLI Commands

    Use SAHI from the command-line for prediction and dataset operations.

    Learn more

Interactive Examples

All documentation files are complemented by interactive Jupyter notebooks in the demo directory: