This app utilizes two object detection models, with the option of adding additional detection models. This may be helpful for including models that comprise very different libraries. In this example, one model detects numerous objects of small to medium size, and the other has a more limited library but detects some larger objects, such as airplanes, trains, and sofas. As there is some overlap between models, for instance they both detect people, you can compare the prediction confidences. Additionally, the output for each model appears in a separate video frame.
To run this app, you will need an alwaysAI account. Please register at https://alwaysai.co/auth?register=true
Easy start up guides can be found following registration. Please see the docs page for more information: https://alwaysai.co/docs/getting_started/introduction.html
The models used in this app are "alwaysai/mobilenet_ssd" and "alwaysai/ssd_inception_v2_coco_2018_01_28". You can find out more about these models at the following pages: (alwaysai/mobilenet_ssd): https://alwaysai.co/model-catalog?model=alwaysai/mobilenet_ssd
(alwaysai/ssd_inception_v2_coco_2018_01_28): https://alwaysai.co/model-catalog?model=alwaysai/ssd_inception_v2_coco_2018_01_28
You can alter the code to use different detection models: https://alwaysai.co/docs/application_development/changing_the_model.html
For convenience, there is a variable called 'colors' included in the code. Should you alter 'models' to include more than two detection models, also include a new list of tuples to be used as the color to mark up the detected objects in the new model. In the included code, each 'list' contains only one tuple, so all detection boxes for a given model will appear with that color.
If you are having trouble connecting to your edge device, use the CLI configure command to reset the device. Please see the following page for more details: https://alwaysai.co/docs/reference/cli_commands.html
You can also post questions and comments on our Discord Community at: https://discord.gg/alwaysai