phraseloceval

KO | EN

gitlite — search

#javascript #python #hacktoberfest #ai #typescript #go #rust #deep-learning #linux #macos #java #awesome #kubernetes #vue

(★ 20)

Phrase Localization Evaluation Toolkit

phraseloceval 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Phrase Localization Evaluation Toolkit

By [Josiah Wang](http://www.josiahwang.com)

### Introduction

This repository contains the libraries and scripts used for evaluating phrase localization in my ICCV 2019 paper (Phrase Localization Without Paired Training Examples).

I found existing evaluation scripts for the phrase localization task difficult to setup and use. I only really needed to evaluate my output, not to run other people's models so that I can evaluate my output! Thus, this toolkit was born!

I wrote it in standard Python, so the script does not need any other dependencies. You only need to provide the script a list of ground truth bounding boxes and a list of predicted bounding boxes, and it will return the accuracy. I hope having a simple and standard evaluation script for the task will make life easier for everyone!

### Using the toolkit

The toolkit is a Python module located in the ``lib/`` directory. Please refer to the doc comments in the code for explanations and usage.

Ground truth annotations for various datasets are provided in the ``data/`` directory.

An example script is available as ``demo.py``.

### Citation

If you use this evaluation toolkit, please cite the following work:

Josiah Wang and Lucia Specia (2019). **Phrase Localization Without Paired Training Examples**. In *Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)*.

@InProceedings{WangSpecia:2019,
	author    = {Wang, Josiah and Specia, Lucia},
	title     = {Phrase Localization Without Paired Training Examples},
	booktitle = {Proceedings of the IEEE/CVF Internaitonal Conference on Computer Vision (ICCV)},
	publisher = {{IEEE}},
	month     = oct,
	year      = {2019},
	pages     = {},  
	address   = {Seoul, South Korea}
	}

### License

GNU General Public License v3.0

phraseloceval

phraseloceval 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

phraseloceval

phraseloceval 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README