Bringing Object Detection to the Masses: Easily Locating Items with a Few Clicks!

Posted by

BigML’s new release is here! Join us on Tuesday, June 28, 2022, at 08:00 AM PDT / 05:00 PM CEST for a FREE live webinar to discover the latest improvements for our Machine Learning platform. We will be announcing Object Detection, a feature that will boost your productivity significantly as you look to generate valuable business insights from your image data. Object Detection is extremely helpful to solve a wide range of computer vision use cases such as medical image analysis, quality control in manufacturing, license plate recognition in transportation, and people detection in security surveillance, among many others!

BigML’s Object Detection release is a continuation of and improvement on our previous Image Processing release. With Object Detection you can easily label image data, locate objects and annotate regions in it, or even import pre-defined regions in other popular formats (pascal-voc, YOLO, among others) and edit them with a handy web interface. Once your image regions are defined, you can train and evaluate Object Detection models, make predictions, and automate end-to-end Machine Learning workflows on a single platform

Object Detection can be used all across the board, in all sorts of industries. To name a few examples, in security it can be used for people detection and counting; in transportation, for detection of vehicles, license plates, traffic signs, traffic lights, crosswalks, and road blocks; in healthcare, Object Detection helps easily find tumors and other abnormalities in radiological images; in retail, it is used for merchandise detection and counting; in manufacturing, it can be leveraged to detect and count defective outputs; in agriculture, Object Detection can be used to find weeds, pests, and unhealthy plants. 

Object Detection goes a step further than our previous release. While Image Processing answers whether a certain class is present in an image, Object Detection does not only answer whether multiple classes are present in an image but is also able to locate them in the image. To make that possible, BigML enables Object Detection by introducing the regions optype. Using it, you can annotate bounding boxes on the platform and seamlessly integrate Object Detection into many Machine Learning workflows.

Of course, all of this functionality is also available from BigML’s API, which means that datasets and models for Object Detection can easily be created with simple REST calls or via the use of the BigML bindings. One can use these same calls to make predictions or download the Object Detection models to make predictions locally! Object Detection is also available from WhizzML to automate your Machine Learning workflows.

BigML’s intuitive visualizations make your life easier by displaying global performance as well as per image scorings. Also, images are grouped by correct or incorrect predictions and provide a detailed view of ground truth and predicted regions in each image. 

Finally, by using the BigML platform for Object Detection, you get all of the functionality that BigML brings to all its Machine Learning capabilities, including evaluations with comprehensive visualizations, batch predictions at scale, resources that are always immutable and traceable, and server-side scripting that enables the execution of complex workflows with a single click.

As usual, BigML’s goal is to let you focus on what really matters: your business problem. This means, among other things, not having to worry about running specialized GPU servers, assuring compatibility between countless software libraries, or micro-managing system resources. BigML eliminates all that complexity by automatically allocating system resources to optimize image processing tasks.

Do you want to know more about Object Detection?

Please visit the dedicated release page to know more about Object Detection and join the FREE live webinar on Tuesday, June 28, 2022, at 08:00 AM PDT / 05:00 PM CEST. Register today, space is limited!

Finally, stay tuned for the upcoming series of blog posts about Object Detection, where we will start by explaining the basic concepts of this new feature and continue showcasing it with example use cases, and tutorials on how to use Object Detection through the BigML Dashboard, API, WhizzML and the Python Bindings, as well as how this new feature has been implemented on the BigML platform.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s