Image & Bounding Box Augmentation using Albumentations | Essentials of Object Detection

This tutorial explains how to do image pre-processing and data augmentation using Albumentations library.
Google Colab notebook:
colab.research.google.com/dri...

[@ssshukla26]

I was waiting for your videos from last several weeks. And there is 3 videos in last 2 days. Christmas came early this year...

[@KapilSachdeva]

You are too kind to me Sunny!

[@dave11a82]

Really well explained. Thanks for this tutorial!

[@pradumnar870]

Very clear and intuitive explanation....please keep doing this kind of tutorials

[@KapilSachdeva]

I will try my best 🙏

[@DSJOfficial94]

this is so clear and good

[@KapilSachdeva]

🙏

[@harshith_takkala]

thanks for this tutorial

[@KapilSachdeva]

🙏

[@abimbolaadeola1986]

Hello. Thank you very much for the tutorial.
But will this method work for my dataset? I have quite a large number of images that need to be resized together with bounding box coordinates in xml formats. Please, any idea?
Thank you very much.

[@yonnileung]

When i install the albumentations in Anaconda environment by "pip install albumentations, it upgrade my numpy to 1.24.4 automatically, this version of numpy conflict to Tensorflow 2.3.0...

[@Game-nr3sr]

In Industry can we use Roboflow for data augmentation. Even though its paid, but is it desirable to use roboflow?

[@KapilSachdeva]

Have not used it myself so can not really give an informed advice.

[@Game-nr3sr]

@@KapilSachdeva ok sir 🙂

[@chaitanyayadav3917]

Can we use this for detecting signature froom an image?

[@KapilSachdeva]

Yes but first you will have to create a dataset of images with signatures with ground truth bounding boxes and train your network.

[@chaitanyayadav3917]

@@KapilSachdeva sure sir , also can you give me a reference link to get more idea or this video is enough ?