OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Developer Tools & SDKs

Samexporter

Samexporter is a Python library and command-line tool that exports supported Segment Anything-family models to ONNX and runs image segmentation with ONNX Runtime.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Samexporter — OpenViST project profile in AI Developer Tools & SDKs.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Samexporter when a project needs prompted segmentation from supported SAM-family models in an ONNX Runtime workflow, or when exporting those models from their original checkpoints.

Who it is for

  • Computer-vision developers preparing SAM-family models for ONNX Runtime.
  • Teams building self-hosted image-segmentation inference workflows.

Documented uses

  • Export supported SAM-family checkpoints to ONNX models.
  • Run point- or rectangle-prompted image segmentation with ONNX Runtime.
  • Run text-prompted segmentation with SAM 3 using its image, language, and decoder ONNX models.
Before you choose

Requirements & limits

The repository is public and identifies its license as MIT. Installation instructions require Python 3.11 or newer and document a CPU runtime installation.

  • Different model families require different ONNX files and may need a different runtime variant; EfficientSAM is explicitly selected with --sam_variant efficient_sam.
  • SAM and MobileSAM export workflows begin with checkpoints, while the EfficientSAM guide uses official pre-existing ONNX files.
  • SAM 3 uses three ONNX models and adds text prompting; its source export setup has additional NumPy and package requirements.
  • The repository's MIT license does not establish the licenses of every upstream model; check the applicable model terms.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Run inference with EfficientSAM-Ti ONNX files

Run point- or rectangle-prompted segmentation using the documented EfficientSAM-Ti runtime workflow.

Before you start

  • Python 3.11 or newer.
  • Samexporter installed with its CPU runtime extra and the documented PyTorch and torchvision versions.
  • EfficientSAM-Ti encoder and decoder ONNX files.
  • An input image and a prompt JSON file.

Steps

  1. Install PyTorch 2.10.0 and torchvision 0.25.0 using the documented CPU wheel index, then install samexporter[runtime-cpu].

  2. Download the EfficientSAM-Ti encoder and decoder ONNX files to output_models/efficient_sam/.

  3. Run the command below with --sam_variant efficient_sam, the encoder and decoder paths, --image, --prompt, and --output arguments.

    Documented command
    python -m samexporter.inference

Expected outcome

The documented command writes a segmentation result to the specified output path for point or rectangle inference.

Guide limitations

  • This workflow uses pre-existing EfficientSAM ONNX files; it does not export a model checkpoint.
  • The example depends on the user supplying an image and a prompt JSON file.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
14 October 2026
Original website title
vietanhdev/samexporter · Official homepage title · retained in its source language
How we review sourcesSuggest a correction