https://github.com/enobufs/node-menoh-yolo

Object classification using Yolo powered by Menoh.
https://github.com/enobufs/node-menoh-yolo

Last synced: 10 days ago
JSON representation

Object classification using Yolo powered by Menoh.

• Host: GitHub
• URL: https://github.com/enobufs/node-menoh-yolo
• Owner: enobufs
• License: mit
• Created: 2018-10-26T18:38:51.000Z (about 6 years ago)
• Default Branch: master
• Last Pushed: 2024-04-10T11:36:29.000Z (7 months ago)
• Last Synced: 2024-04-14T06:01:41.348Z (7 months ago)
• Language: JavaScript
• Size: 4.24 MB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# menoh-yolo

[![npm version](https://badge.fury.io/js/menoh-yolo.svg)](https://badge.fury.io/js/menoh-yolo)

[![Build Status](https://travis-ci.org/enobufs/node-menoh-yolo.svg?branch=master)](https://travis-ci.org/enobufs/node-menoh-yolo)

[![Coverage Status](https://coveralls.io/repos/github/enobufs/node-menoh-yolo/badge.svg?branch=master)](https://coveralls.io/github/enobufs/node-menoh-yolo?branch=master)

Object classification using YOLO v2 powered by [Menoh](https://github.com/pfnet-research/menoh).

This module is made on top of npm module, [menoh](https://github.com/pfnet-research/node-menoh), which is a wrapper of Menoh (C/C++) core.

The modules supports both command line interface and API.

## Requirements

You must have the following libraries installed in advance.

* MKL_DNN Library (0.14 or later)

* Protocol Buffers (2.6.1 or later)

* Menoh (core (1.1.1 or later)

> See [Building Menoh](https://github.com/pfnet-research/menoh/blob/v1.1.1/BUILDING.md) for more

> details.

## Installation (command line use)

```

npm install -g menoh-yolo

```

## Usage

```

$ menoh-yolo --help

Usage: menoh-yolo [options]

Options:

  -v, --version            Output the version number

  -i, --input    Input file path.

  -o, --output   Output file path.

  -s, --score      Score threshold (0, 1.0] (defaults to 0.4)

  -x, --overlap    Overlap threshold (0, 1.0] (defaults to 0.5)

  -b, --browse             Show output image with default viewer.

  -h, --help               Output usage information

```

### Example

```

$ menoh-yolo -i ./dog.jpg -b

```

> When you use the CLI for the first time, it automatically downloads ONNX file and a

> configuration file into $HOME/.menoh-yolo/.

By giving `-b` option, it will open up a viewer to show the output result like below:

![Alt text](./doc/output.png?raw=true "Output Image")

## API

(TODO)

> See `./bin/cli.js` as an example.