YOLO(You Only Look Once):Real-Time Object Detection

Caffe-YOLO实现

最新推荐文章于 2024-11-05 22:52:15 发布

weixin_34189116

最新推荐文章于 2024-11-05 22:52:15 发布

阅读量114

点赞数

CC 4.0 BY-SA版权

文章标签： python 人工智能

原文链接：http://www.cnblogs.com/ystwyfe/p/7942953.html

本文介绍了一个基于Caffe框架的YOLO实时目标检测系统的实现方法。该项目支持YOLOv1，并通过四个步骤将Darknet预训练权重转换为Caffe模型。首先从Darknet的配置文件创建Caffe的prototxt文件，初始化Caffe网络和权重，读取Darknet的预训练权重文件，并用这些权重替换Caffe中初始化的权重。

caffe-yolo:https://github.com/xingwangsfu/caffe-yolo

YOLO in caffe

Update 12-05-2016: Currently, only yolo v1 (http://pjreddie.com/darknet/yolov1/) is supported. Yolo V2 (http://pjreddie.com/darknet/yolo/) is not supported. Batch norm layer is supported.

Introduction

This is a caffe implementation of the YOLO:Real-Time Object Detection

Note, the caffe models are not trained in caffe, but converted from darknet's (.weight) files (http://pjreddie.com/darknet/yolov1/).

The converter is consisted of four steps:

create .prototxt files for caffe from .cfg file in darkent
initialize the caffe network and weights from .prototxt file
read the weights from pre-trained darknet's (.weight) file
replace the initialized weights with the weights in pre-trained darkenet's (.weight) file

Usage

run "create_yolo_prototxt.py" to create .prototxt files
run "create_yolo_caffemodel.py" to create .caffemodel files
run "yolo_main.py" to call yolo in caffe

Convert yolo's (.weight) files to caffemodel

first, you need to download the pretrained yolo weight files (http://pjreddie.com/darknet/yolov1/) and .cfg files (https://github.com/pjreddie/darknet/tree/master/cfg/yolov1)
run create_yolo_prototxt.py to create .prototxt files
after that, run create_yolo_caffemodel.py to create the caffemodel from yolo's (.weight) files
- "python create_yolo_caffemodel.py -m train_val_prototxt.filename -w yoloweights_filename -o caffemodel_filename"
replace train_val_prototxt.filename with /your/path/to/yolo_train_val.prototxt (yolo_small, yolo_tiny), yoloweights_filename with /your/path/to/yolo.weights (yolo-small, yolo-tiny), and caffemodel_filename with your output caffemodel name,

e.g. "python create_yolo_caffemodel.py -m yolo_train_val.prototxt -w yolo.weights -o yolo.caffemodel"

Main file usage

run yolo_main.py to do yolo object detection for the input image

"python yolo_main.py -m model_filename -w weight_filename -i image_filename"

replace model_filename with /your/path/to/yolo_small_deploy.prototxt or yolo_tiny_deploy.prototxt, weight_filename with /your/path/to/yolo_tiny.caffemodel or yolo_small.caffemodel and image_filename with the target image file