REALTIME FEATURE DETECTION IN A BROWSER WITH TENSORFLOW JS

At Deductive, we've been doing AI and machine learning for a long time. When we started, it was hard, requiring massive computing power and a lot of time. At the beginning of 2019, we are in a situation where many of these technologies are mature to the stage where they will run in your browser in JavaScript. The demo below runs the TensorFlow COCO SSD model in your browser to detect in realtime features from the COCO dataset. It should identify everyday objects like people, cups, cats, cell phones, laptops, etc...

Click the button below to enable your webcam and download the model (26Mb in size).

If you need help productizing any of the open source machine learning models, get in touch with us using the form below.

Thanks go to Nick Bourdakos whose article is the inspiration for this page.