alturosdestinations / alturos.imageannotation Goto Github PK

A collaborative tool for labeling image data for yolo

License: MIT License

C# 97.83% PowerShell 2.17%

annotation-tool labelling labeling labeling-tool tagging taggingtools image-tagging image-annotation bounding-boxes boundingbox

alturos.imageannotation's Introduction

Alturos.ImageAnnotation

The purpose of this project is to manage training data for Neural Networks. The images are stored in an object storage for example Amazon S3. In our case we have image data for different runs that we want to annotate together. You can upload a folder into a package. For every package you can set your own tags... this information is stored in a database for example Amazon DynamoDB.

Features

Collaborative annotation of images
Verification of image annotation data
Export for yolo (train.txt, test.txt, obj.names) with filters
No requirement for a custom server

Installation

You have the choice to use AWS or install your own environment.

Keyboard Shortcuts

Shortcut	Description
`↓`	Next image
`↑`	Previous image
`→`	Next Object Class
`←`	Previous Object Class
`0`-`9`	Select Object Class
`WASD` +`Shift` +`Ctrl` +`Alt`	Move Bounding Box Resize Quick Invert

Data preperation

How many images are required

You should preferably have 2000 different images for each class or more.

Extract images from a video

If you have a video file and need the individual frames you can use ffmpeg to extract the images. This command exports every 10th frame in the video. ffmpeg -i input.mp4 -vf "select=not(mod(n\,10))" -vsync vfr 1_every_10/img_%03d.jpg

Articles of interest

Training YOLOv3 : Deep Learning based Custom Object Detector

Credits

This program uses icons from the Silk icon set created by Mark James, which can be found here. The icon set is licensed under a CC BY 3.0 license. Some changes were made to the icons.

Other Image Annotation Tools

List of annotation tools for machine learning research

alturos.imageannotation's People

Contributors

Stargazers

Watchers

Forkers

lkneringer innozent monkeybin11 anuragsinghchaudhary get-resource shalevy1 deezy69 reddevil1310 tinohager xiaodelea futureflsl eighteen-k-gold-malow nguyenanhtuan1008 geraldoarema mlnethub webstorage119 chai104 877325778

alturos.imageannotation's Issues

Cannot install locally

Hi, I am getting this error when trying to use your script, why is that?

Fix Hotkeys for resizing clashing with UI

Modifier keys don't work well at the moment because they trigger functions in the UI instead of being restricted to manipulating the annotation border

Errors, bugs

Actually I dont get it how to even assign labels to the image.. At first application compiled and opened, I was able to add some tags to the list, but could not assign it to image, then i tried to add some objects because the only label on the image I could see was default "MyObject" after doing so application does not compile anymore... throwing this exception, could you please add more detailed information how to actuality assign a label to an image? Also app seems to only accept jpg type, when selecting png it return white blank image...

Add key to prevent other boxes from being selected

Sometimes a border can't be selected well because the anchor point you want to grab is unavailable due to another border overlaying it. Holding down a certain key should lock the selection to the currently highlighted object so it can be grabbed easily.

how to convert output to trained dataset

Hi Tino, i used the annotations example provided and labelled the image in the package. u got the output as train.txt,train.cfg and obj.names.
My concern is that how to convert it to a valid trained dataset for the object detection with yolo.alturos.
I need it in the form of .cfg,.weight and .txt so that I can load it in my object detection application.

It would be a great help if you let me know the procedure for the same. Thanks in advance.

Add package didn't upload package to Amazon S3

@tinohager Thank you so much for your hard work.
I'm trying to add a package 'test' by Alturos.ImageAnnotation1.0.6 but I saw that the package didn't upload to Amazon S3.
It only saves into my local folder: Alturos.ImageAnnotation1.0.6\packages\test.
Did I set any wrong in my DynamoDB or S3? My config DynamoDB like below:

Is my primary partition key, and Primary sort key correct? How do you set it when you create the DynamoDB table?

And after I add package 'test', nothing uploads into my S3 bucket. What happening?

This is my IAM policy JSON:

Looking forward to your reply.
Thank you so much.

Create a report about object classes before export

total number of annotated objects of one class

DirectoryNotFoundException: Could not find a part of the path for cfg files

Hi I got an error when I try to export the data.

DirectoryNotFoundException: Could not find a part of the path 'C:\Users*****\source\Alturos.ImageAnnotation-master\Alturos.ImageAnnotation-master\src\Alturos.ImageAnnotation\bin\Resources\yolov3.cfg'.

I tried going to the folder and the file is there and I don't see any problem :/
Please help