Hey team! I’m currently working on a project where I get a bunch of image data with bounding box labels. Unfortunately, the dataset quality isn’t perfect and I need to go through the dataset and manually filter out the images with bad labels. I spent ~half a day building a little annotation tool with matplotlib last week, but the results really weren’t pretty. Today, I decided to do some more comprehensive googling first(literally the first link of my first search) and found this really cool annotation library called pigeon that does the job pretty well and the code to use it is dead simple. If anyone else has a labeling library that they use, I’d love to hear about it!
1 Like
I added a back button to the library along with an example of how to use it with fastai’s Image class.
Getting it to work with fastai’s Image.show method was a little tricky, but I was able to make it work with a minor hack. Suggestions for improvement/library alternatives are hugely welcome!
https://github.com/csaroff/pigeon/blob/master/examples/back_button.ipynb