thtrieu/darkflow

How to save features of each detected bounding box?

Open

#65 创建于 2017年2月22日

在 GitHub 查看
 (3 评论) (4 反应) (0 负责人)Python (6,120 star) (2,102 fork)batch import
help wanted

描述

I would like to save the features of each detected bounding box for later use. How can this be done? How can I select the specific layer from which I want the features to be extracted? Thank you

贡献者指南

技术栈
pythontensorflow
领域
machine learningai
议题类型
research
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。
under 1 hour
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
basic PythonTensorFlow basicsobject detection concepts
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。
40
研究方向
To extract features from a specific layer, you need to modify the inference script (e.g., darkflow/net/flow.py) to intercept the output of a given layer. Identify the layer name from the YOLO architecture (e.g., 'conv 23') and add a TensorFlow hook using tf.Graph.get tensor by name(). Then, feed the input image and retrieve the tensor values for each bounding box. Check existing comments in the issue for hints on implementation.