NaturalIntelligence/imglab

Exporting data in Tensor flow format

Open

#84 创建于 2018年8月21日

在 GitHub 查看
 (1 评论) (0 反应) (0 负责人)HTML (930 star) (560 fork)batch import
enhancementgood first issuehelp wanted

描述

Currently imglab supports exporting labelled data in dlib, Pascal VOC, and COCO formats. We want to let users export data in tfrecord (tensor flow) format as well.

target folder : data formaters

Please check how export feature is implemented for other formats in data formaters

Some code reference to generate tfrecord: https://github.com/datitran/raccoon_dataset

Watch for changes, or Bookmark for easy discovery. Fund this project for new features and maintenance. [Showcase] your project with us by raising an issue

Want to back this issue? Post a bounty on it! We accept bounties via Bountysource.

贡献者指南

技术栈
pythontensorflow
领域
data
议题类型
feature
难度面向新贡献者的预计实现难度，1 表示很小改动，5 表示专家级工作。
3
预计时间有经验贡献者完成调查、实现、测试并准备 pull request 的粗略时间范围。
1-2 days
活动状态议题当前的可参与程度：新鲜、活跃、陈旧、阻塞或等待维护者输入。
stale
清晰度议题是否清楚说明期望改动、验收标准和下一步。
clear
前置要求
Knowledge of PythonFamiliarity with TFRecord formatUnderstanding of imglab's existing export architecture
新手友好度1-100 的估计分数，表示该议题对首次贡献者的友好程度。
60
研究方向
Start by examining the existing export implementations in the 'dataformaters' directory (linked in the issue). Understand how the current export formats (dlib, Pascal VOC, COCO) are structured and integrated. Then, use the provided raccoon dataset repository as a reference for generating TFRecord files. The goal is to create a new exporter module that outputs TFRecord format, following the same pattern as other exporters.