Run neural network applications at 5.0 TOPS with edge-AI SOC as open-root, with advantages for deciding the bootloader’s booting options, and managing team members for programming. Benefit from Open-root development for code transparency, no hidden background process, and freedom to incorporate libraries based on need.

Open-root with Edge Computing

Run neural network applications at 5.0 TOPS with edge-AI SOC as open-root. (Note: Please consult for the availability of the Development Kit)

Own your code, with Open-Root

Take advantage of Open-Root Mode

Decide your own programs
Code transparency within the programmer’s code
No hidden or unclear background process running
Customization and freedom to incorporate libraries based on need

The Open-Root Mode Architecture

Within an embedded environment product, having a minimal file system available is widely preferred to enhance further developments. Installing the complete operative system and the complete package of dependencies can be troublesome as many files may not be needed for the specific application use-case.

The Solution:

Access as Root mode (i.e.: to the roof filesystem) for a developer can offer flexibility to customize the basic operating system into the minimal and incorporating further packages per necessity offers limitless development options.

Power Efficiency & Peak Performance

Efficient for Edge Computing

Powered by Quad core ARM Cortex-A73 and dual core ARM Cortex-A53 CPU
Mali GPU G52 MP4 at 850Mhz
Internal L2 cache (512KB) and system workspace buffer (1MB)

Superior Performance

5.0 TOPS Performance
Superior CPU & GPU performance
Expandability for Embedded Solutions

Object Detection with YOLO & Deep Learning with TensorFlow™ and Caffe™

Develop and run the most widely used open-source software library for machine learning and artificial intelligence

YOLO (You Only Look Once)

Object Detection
Open Source Codes

TensorFlow™

Focus on training and inference of deep neural networks
Supported languages: including Python, JavaScript, C++, and Java.

Note: supported by Amlogic

Caffe™

A framework that supports many different types of deep learning architectures geared towards image classification and image segmentation
CNN, RCNN, LSTM and fully connected neural network designs.

Note: supported by Amlogic

Note:
"TensorFlow, the TensorFlow logo and any related marks are trademarks of Google Inc.”
“Caffe is released under the BSD 2-Clause license. The BAIR/BVLC reference models are released for unrestricted use.”

For transcoding TensorFlow and Caffe, material are available through Github, from this link

Easy Access with Remote Desktop Protocol

Secure access using RDP, enabling developers to remotely diagnose problems from physical work desktop computers

Both Windows and Linux terminals

Easy Debugging through UART

Access and debug simply through UART console

An IMU inside

The LIPSedge ™ AE430-DK & AE470-DK comes with a built-in 6 axis Inertial Measurement Unit (IMU), which allows precision and accuracy for its position and movement location details.

Persistent Connection

Supports GigE with PoE that allows data transmission and power supply (802.3 af / at) up to 100 meters
Remain in the same session even after occasional disconnection (<15 seconds) without restarting

Know more…

Build Once, Deploy Unlimitedly

Cross-platform OS support on Windows & Ubuntu with common SDK. Android & MacOS support is upon request
Cross-platform hardware support on Intel (x86) and Nvidia Jetson (arm)

Develop with LIPSedge™ SDK

LIPSedge SDK™ supports variety of programming languages including Python, C, C++, C#, Visual Basic, Java, Flutter, and more
LIPSedge SDK™ also supports widely used industrial 3D framework including OpenNI, OpenCV, ROS/ROS2, Nvidia Isaac, Halcon.

Strong LIPSense™ Middleware Support

LIPSedge™ AE430 is compatible with LIPSense™ middleware including LIPSense™ 3D Body Pose SDK, LIPScan™ 3D SDK, and more

Application uses cases

vision guided robotics

LIPScan 3D Scan SDK

LIPSafeGuard 3D Virtual Fence

Start developing
with LIPS

Documentation

User Manual

Specifications

Depth

Technology:

Active Stereo

Ideal Range:

4 meters

Frame Rate:

Up to 60fps

FoV (H x V x D):

65° x 40° x 72°

XY Resolution:

1280x720@30fps, 640x480@60fps

Min. Working Distance:

0.16m

Shutter Type:

Rolling Shutter

Z Accuracy:

≤2% (up to 2 meters and 80% of FoV)

*The Accuracy will vary with disance.

RGB

Resolution:

1920x1080@15fps,
1280x720@30fps,
640x480@60fps

FoV (H x V x D):

69.4° x 42.5° x 77° ( ±3°)

Illumination

Type:

Infared

Wavelength:

850nm

Inertial Measurement Unit

IMU Sensor:

6-axis (3D Accelerator & Gyroscope)

A.I. Processor

ARM Cortex:

Quad Core ARM Corex-A73, paired with Dual Core Cortex A-53 CPU, ARM Mali-G52 GPU

Quick Debug

UART Console:

Rx, Tx, GND

General

Dimension (mm):

118 x 60 x 45 mm

Storage Temperature:

-40 ~ 60˚C

Built-In:

Industrial Heatsink

Power:

PoE (IEEE 802.3 at/at)

M12-to-RJ45 Cable (included)

Power wire (12V, 2A) with M8 standard connector

Weight(g):

Approximately 280 g

Ambient Temperature:

0 ~ 40˚C

Working Environment:

Indoor / Outdoor

Output Interface:

M12 to Gigabit Ethernet cable for power and data

OS(for QEMU setup)

Windows 10

Ubuntu 18.04 / 20.04 / 22.04 LTS

Depth

Technology:

Active Stereo

Ideal Range:

4 meters

Frame Rate:

1280x720@30fps

FoV (H x V x D):

87° x 58° x 95°

Baseline:

95mm

Min. Working Distance:

0.52m

Shutter Type:

Global Shutter

Z Accuracy:

≤2% (up to 4 meters and 80% of FoV)

*The Accuracy will vary with disance.

RGB

Resolution:

1280x800@30fps

Shutter Type:

Global Shutter

FoV (H x V x D):

90° x 65° x 98°

Illumination

Type:

Infared

Pattern Type:

Static

Wavelength:

850nm

Illumating Component:

Vertical-cavity surface-emitting laser (VCSEL)

A.I. Processor

ARM Cortex:

Quad Core ARM Corex-A73, paired with Dual Core Cortex A-53 CPU, ARM Mali-G52 GPU

Quick Debug

UART Console:

Rx, Tx, GND

General

Dimension (mm):

130 x 66 x 42 mm

Storage Temperature:

-40 ~ 70˚C

Built-In:

Industrial Heatsink

Power:

PoE (IEEE 802.3 at/at)

M12-to-RJ45 Cable (included)

Power wire (12V, 2A) with M8 standard connector

Weight(g):

320 g

Ambient Temperature:

0 ~ 40˚C

Working Environment:

Indoor / Outdoor

Output Interface:

M12 to Gigabit Ethernet cable for power and data

OS (for QEMU setup)

Windows 10

Ubuntu 18.04 / 20.04 / 22.04 LTS

Get it Today

LIPSedge™ AE430-DK & AE470-DK is currently on promotion now. Click link below to get your sample today!

LIPSedge™ AE430-DK & LIPSedge™ AE470-DK

​Open-root Development Kit with LIPSedge™ AE Series

Edge Computing made it easy for 3D Stereo Camera

Open-root with Edge Computing

Own your code, with Open-Root

The Open-Root Mode Architecture

Power Efficiency & Peak Performance

Efficient for Edge Computing

Superior Performance

Object Detection with YOLO & Deep Learning with TensorFlow™ and Caffe™

YOLO (You Only Look Once)

TensorFlow™

Caffe™

Note: "TensorFlow, the TensorFlow logo and any related marks are trademarks of Google Inc.” “Caffe is released under the BSD 2-Clause license. The BAIR/BVLC reference models are released for unrestricted use.”

Easy Access with Remote Desktop Protocol

Easy Debugging through UART

An IMU inside

Persistent Connection

Build Once, Deploy Unlimitedly

Develop with LIPSedge™ SDK

Strong LIPSense™ Middleware Support

Application uses cases

vision guided robotics

LIPScan 3D Scan SDK

LIPSafeGuard 3D Virtual Fence

Start developing with LIPS

Documentation

Specifications

Depth

Technology:

Active Stereo

Ideal Range:

4 meters

Frame Rate:

Up to 60fps

FoV (H x V x D):

65° x 40° x 72°

XY Resolution:

1280x720@30fps, 640x480@60fps

Min. Working Distance:

0.16m

Shutter Type:

Rolling Shutter

Z Accuracy:

≤2% (up to 2 meters and 80% of FoV)

*The Accuracy will vary with disance.

RGB

Resolution:

1920x1080@15fps, 1280x720@30fps, 640x480@60fps

FoV (H x V x D):

69.4° x 42.5° x 77° ( ±3°)

Illumination

Type:

Infared

Wavelength:

850nm

Inertial Measurement Unit

IMU Sensor:

6-axis (3D Accelerator & Gyroscope)

A.I. Processor

ARM Cortex:

​Quad Core ARM Corex-A73, paired with Dual Core Cortex A-53 CPU, ARM Mali-G52 GPU

Quick Debug

UART Console:

Rx, Tx, GND

General

Dimension (mm):

118 x 60 x 45 mm

Storage Temperature:

-40 ~ 60˚C

Built-In:

Industrial Heatsink

Power:

PoE (IEEE 802.3 at/at)

M12-to-RJ45 Cable (included)

Power wire (12V, 2A) with M8 standard connector

Weight(g):

Approximately 280 g

Ambient Temperature:

0 ~ 40˚C

LIPSedge™ AE430-DK &
LIPSedge™ AE470-DK

Open-root Development Kit with LIPSedge™ AE Series

Note:
"TensorFlow, the TensorFlow logo and any related marks are trademarks of Google Inc.”
“Caffe is released under the BSD 2-Clause license. The BAIR/BVLC reference models are released for unrestricted use.”

Start developing
with LIPS

1920x1080@15fps,
1280x720@30fps,
640x480@60fps

Quad Core ARM Corex-A73, paired with Dual Core Cortex A-53 CPU, ARM Mali-G52 GPU