A improvement board to shortly prototype on-device ML merchandise. Scale from prototype to manufacturing with a detachable system-on-module (SoM).
A improvement board to shortly prototype on-device ML merchandise. Scale from prototype to manufacturing with a detachable system-on-module (som)
Performs high-speed ML inferencing: the on-board edge TPU Coprocessor is able to performing 4 trillion operations (tera-operations) per second (tops), utilizing 0.5 watts for every tops (2 tops per watt). For instance, it may well execute state-of-the-art cellular imaginative and prescient fashions reminiscent of mobilenet V2 AT 400 FPS, in an influence environment friendly method
Gives a whole system: a Single-board pc with SoC + ML + wi-fi connectivity, all on the board working a spinoff of Debian Linux We name Mendel, so you may run your favourite Linux instruments with this board
Helps tensorflow Lite: no must construct fashions from the bottom up. Tensorflow Lite fashions may be compiled to run on the sting TPE
Helps automl imaginative and prescient edge: simply construct and deploy Quick, high-accuracy customized picture Classification fashions to your system with automl imaginative and prescient edge
Scale from prototype to manufacturing: considers your manufacturing wants. The som may be faraway from the baseboard, ordered in Bulk, and built-in into your {hardware}
Cpu: NXP I.Mx 8M SoC (Quad Cortex-A53, cortex-m4f)
