This trainer is one of several features we've been developing to make implementation of vision-based AI super easy. The entire DepthAI API allows users to leverage the Intel Myriad X SDK for use with DepthAI and MegaAI hardware.
We've also made most of our hardware open source (github.com/luxonis) so that end users can develop with our Myriad X modules for integration into their specific devices.
The DepthAI hardware we've made can run "spatial AI", which combines hardware-level passive stereo depth along with inference to generate metadata which describes not only what's detected, but where it's detected.
Our latest module, MegaAI, combines a 12MP camera with the Myriad X in form factor half the size of credit card. The hardware features (https://www.crowdsupply.com/luxonis/megaai#hardware-features) include 4k60 max stream rate, hardware image and video encoding, 2.5W max power consumption, and of course 4TOPs inference power. All power and data is through a single USB cable.
1 comment
[ 1.4 ms ] story [ 16.7 ms ] threadWe've also made most of our hardware open source (github.com/luxonis) so that end users can develop with our Myriad X modules for integration into their specific devices.
The DepthAI hardware we've made can run "spatial AI", which combines hardware-level passive stereo depth along with inference to generate metadata which describes not only what's detected, but where it's detected.
Our latest module, MegaAI, combines a 12MP camera with the Myriad X in form factor half the size of credit card. The hardware features (https://www.crowdsupply.com/luxonis/megaai#hardware-features) include 4k60 max stream rate, hardware image and video encoding, 2.5W max power consumption, and of course 4TOPs inference power. All power and data is through a single USB cable.