Object Detector

This article introduces an app that uses Machine Studying with TensorFlow for object detection and classification. The app has two modes: detection mode and classification mode, with the former detecting 80 classes from the COCO dataset with bounding box and the latter identifying 1000 classes from the ImageNet dataset without bounding box. To start and stop live preview, users can simply touch the preview screen. The app uses the TensorFlow Deep Studying software library and 2 MobileNet models for detection and classification, and it can run on Android Slimgs (Developer Preview 6.1). It's worth noting that this app is just a demo of using Machine Studying with TensorFlow on Android devices, and no internet connection is required to use it.

Help me rewrite the following article briefly and list its characteristics！Object detection and classification by using Machine Learning with TensorFlow*This app has 2 modes:- Detection mode: detection of 80 classes from COCO dataset with bounding box.- Classification mode: identifying of 1000 classes from ImageNet dataset without bounding box.*To start and stop live preview, just touch on the preview screen.*The Deep Learning software library TensorFlow and 2 MobileNet models are used for detection and classification.*This app can also run on Android Things (Developer Preview 6.1).*This app is just a demo of using Machine Learning with Tenso.....

Embark on your journey to the top of the leaderboards.

Characteristics:

• The article introduces an app that uses Machine Studying with TensorFlow for object detection and classification.
• The app has two modes: detection mode and classification mode, with the former detecting 80 classes from the COCO dataset with bounding box and the latter identifying 1000 classes from the ImageNet dataset without bounding box.
• The app uses the TensorFlow Deep Studying software library and 2 MobileNet models for detection and classification.
• The app can run on Android Slimgs (Developer Preview 6.1).
• No internet connection is required to use the app.
• The article is concise and to the marker.

What's New in the Latest Version 1.2.2

Last updated on Mar 26, 2018