{"id":1744,"date":"2019-09-28T11:56:11","date_gmt":"2019-09-28T11:56:11","guid":{"rendered":"http:\/\/uni.hi.is\/helmut\/?p=1744"},"modified":"2019-10-07T10:14:53","modified_gmt":"2019-10-07T10:14:53","slug":"european-researchers-night-from-the-next-generation-supercomputer-deep-est-to-your-smartphone-real-time-object-detection-using-neural-network-2","status":"publish","type":"post","link":"https:\/\/uni.hi.is\/helmut\/2019\/09\/28\/european-researchers-night-from-the-next-generation-supercomputer-deep-est-to-your-smartphone-real-time-object-detection-using-neural-network-2\/","title":{"rendered":"European Researcher's Night: From the next generation supercomputer DEEP-EST to your smartphone -- real-time object detection using neural network"},"content":{"rendered":"

The DEEP-EST research project<\/a> is at V\u00edsindavaka, part of the European Researcher's Night<\/a>, in Reykjavik, 28. September 2019.<\/p>\n

Use the camera of your smartphone to detect objects in real-time. While neural networks are still best trained on a supercomputer, such as DEEP-EST with its Data Analysis Module<\/a>, the trained neural network even runs in the browser of a smartphone. Bring your smartphone and objects such as apples, bananas or teddy bears to let your smartphone detect these objects.<\/p>\n

$\"https:\/\/nvndr.csb.app\/\"$ <\/p>\n

Just open the following web page and allow your browser to use the camera: https:\/\/nvndr.csb.app\/<\/a>.<\/p>\n

(Allow a few seconds for loading the trained model and initialisation.)<\/p>\n

The used approach is Single Shot Detector (SSD)<\/a> (the percentage shows how sure the neural network is about the classification) using the Mobilenet neural network architecture. The dataset used for training is COCO (Common Objects in Context), i.e. only objects of the labeled object classes contained in COCO<\/a> will get detected. The Javascript code that is running in your browser<\/a> uses Tensorflow Lite<\/a> and its Object Detection API<\/a> and model zoo<\/a>.<\/p>\n

If you want learn more about DEEP-EST, have a look at the poster below (click for PDF version):<\/p>\n

$\"PDF$ <\/a><\/p>\n","protected":false},"excerpt":{"rendered":"