Deep-ET
KOContact Deep-ET

Product

Product

Compare all five product families — deployment targets, supported boards, and AI applications — in one place.

YBPU on Edge

An AI deployment framework for low-cost embedded and IoT boards such as Nvidia Jetson, Rockchip, and Raspberry Pi.

Nvidia Jetson SeriesRockchip 3328/3399Raspberry Pi 3/4

Applications

Edge AI Application

Re-ID

Tracks an identified person continuously in real time on embedded hardware using a tracking model fed by live camera input.

Customer Tracking in Building

Segmentation

Runs segmentation on embedded and IoT boards to distinguish objects in real time, including instance-level separation of similar objects.

Image Segmentation

YBPU on Mobile

An AI framework for Android and iOS environments built around mobile chipsets such as Qualcomm Snapdragon and Apple Bionic.

Qualcomm Snapdragon 8xxApple Bionic

Applications

Mobile AI Application

Segmentation

Runs real-time segmentation on mobile hardware, including instance segmentation to separate visually similar objects.

Image Segmentation

Deep ET Face ID

Determines in real time whether a face captured on mobile is live or fake, including printed photos, screens, silicone masks, and 3D-rendered faces.

3D Face DetectionFace Liveness Detection

YBPU on System

An AI framework for laptops and low-cost systems using integrated-GPU notebooks, AMD Ryzen 4000 series devices, and Intel i3-class CPUs.

AMD GPUNvidia GPUNvidia A100/V100AMD CPUIntel CPU

Applications

Laptop AI Application

Quad AI Detection

Runs four deep-learning applications at the same time on low-cost CPU notebooks, including object and face recognition workloads.

Quad AI Detection

YBPU on IoT

An AI framework for low-cost IoT boards such as Arduino M0 and RISC-V devices.

Risc-VArduino M0Arduino Nano 33 BLE

Applications

IoT AI Application

Object Detection

Ports object-detection models to IoT boards so real-time camera streams can produce size, position, and direction data.

2D Object Detection

YBNW on NPU

YBNW is a YB NPU Wrapper that integrates multiple NPUs behind Python and C interfaces so trained models can be ported to target NPUs in a single step.

ITX-3588J(Firefly)Dev Board(Coral)Vim3(KHADAS)

Applications

YBNW Application

Object Detection

A cross-platform NPU wrapper used to port modern object-detection models and maximize NPU resources for higher throughput.

2D Object Detection

X-ray Detection

A cross-platform NPU wrapper used to port X-ray detection models and accelerate inference beyond conventional object-detection engines.

X-ray Detection