IBASE

Embedded Computing

SI-62S-E / SI-62S-S

AI Education Platform - Powered by 9th/8th Gen Intel® Core™ Desktop Processor-based with 1x PCI-E (x16) or 2x PCI-E (x8) Slots

A comprehensive AI deep learning and edge computing solution designed for educational providersDesigned for collecting and analyzing student's behavior and engagement in the classroom through facial recognitionOptimize teaching quality through classroom interaction analysisFacial Recognition - 99% face verification accuracyBiometric identification - 68 points 3D landmarks detection for 6 kinds of student face emotion identificationBehavior identification - 17 human key-points (skeleton) detection for 9 kinds of student behavior identification
Specification

The IBASE AI Education Platform is a comprehensive AI deep learning and edge computing solution for educational providers that collects and analyzes student's behavior and engagement in the classroom using facial recognition. The solution combines the IBASE SI-62S Education system, GEN3G-200 FHD camera and the YUAN SC550N4 SDI capture card with Intel OpenVINO based AI QDEEP SDK development tool for facial recognition. Facial recognition AI can interpret student data using quantitative methods to help educators to provide efficient teaching in educational environments. With the aid of facial recognition, students' behavior and class engagement become quantifiable data that can be used to promote effective teaching in the educational environment. 


The platform uses AI image analysis with facial recognition to monitor and analyze students' behavior and performance in classrooms. It aims to provide real-time data to the teachers and supervisors of the students’ emotions exhibited. The technology classifies the students based on the range of emotions displayed – from antipathy to joy and engagement. It is programmed to record various actions and situations such as writing, reading, raising of hands, and even falling asleep. This not only helps the faculty to identify inattentive students, it also signals teachers to alter their teaching style in order to get a more positive result.

The AI QDEEP SDK included in this solution applies the Human Key point Detection approach in computer vision. By quickly locating and returning the 3D coordinates of each human key point, AI QDEEP SDK precisely detects human skeleton, behavioral modes and motions. Furthermore, behavior detection and analysis can be applied to smart classrooms. By reviewing the collected data on student performance, lecturers can make the needed adjustments in instructions and lesson content based on the students' progress.

 

IBASE AI Education Platform

System Mainboard MI998AF
CPU Type 9th/8th Generation Intel® desktop processors TDP<=65W
CPU Socket LGA1151
Chipset Q370 PCH
Memory 2x DDR4 SO-DIMM 2666/2400 MHz , Dual channel, Max.32GB
Graphics Intel® UHD Graphics P630
LAN 1x Intel® I219LM
1x Intel® I211AT
I/O Interface 1x HDMI, 1x DVI-D and 1x DP from MI998 MB
6x USB 3.0, 2x USB 2.0
1x GPIO port
2x RJ45 for Gigabit LAN
2x COM#1/2 port for RS232/422/485
2x COM#3/4 port for RS232 only (Via onboard box-headers)
3x Microjack audio connectors for Mic in, Line-in and
Line-out
1x HDMI in for the capture card option
Power / HDD LED, 1x power button
1x Terminal block for external power button
1x Reset button, 1x DC 12V output
Auto Control and Monitoring 1. 2. ODM/OEM is available. 3. Features
● Behavior identification
● Face Recognition
● Skeleton Detection
● Not limited by number of people
● 1x PCI-E (x16) or 2x PCI-E (x8) based on PCI-E x16
connectors for GPU card options
Specifications
SI-62S System Specifications
System
Mainboard
MI998AF
CPU Type
9th/8th Generation Intel® desktop processors
TDP<=65W
CPU Socket LGA1151
Chipset Q370 PCH
Memory
2x DDR4 SO-DIMM 2666/2400 MHz , Dual channel, Max.
32GB
Graphics Intel® UHD Graphics P630
LAN
1x Intel® I219LM
1x Intel® I211AT
Expansion Slots
1x PCI-E (x16) or 2x PCI-E (x8) based on PCI-E x16
connectors for Matrox / AMD / NVIDIA graphics card
1x M.2 M2280 (NVMe supported)
1x M.2 E2230
I/O Interface
1x HDMI, 1x DVI-D and 1x DP from MI998 MB
6x USB 3.0, 2x USB 2.0
1x GPIO port
2x RJ45 for Gigabit LAN
2x COM#1/2 port for RS232/422/485
2x COM#3/4 port for RS232 only (Via onboard box-headers)
3x Microjack audio connectors for Mic in, Line-in and
Line-out
1x HDMI in for the capture card option
Power / HDD LED, 1x power button
1x Terminal block for external power button
1x Reset button, 1x DC 12V output
Auto Control and
Monitoring
Power Requirement AC 110V~240V
Construction Aluminum + SGCC
Weight 8kgs (17.64Ibs)
Chassis Color Black & white
Storage 2x 2.5” SATA SSD with RAID 1 support
Expansion Slots 1x PCI-E (x16) or 2x PCI-E (x8) based on PCI-E x16
connectors for Matrox / AMD / NVIDIA graphics card
1x M.2 M2280 (NVMe supported)
1x M.2 E2230
Power Supply Internal 500W PSU
Mounting Standard system bracket
Dimensions 406mm(W) x 345mm(D) x 93mm(H)
17.16”(W) x 15.9”(D) x 3.66”(H)
Storage Temperature 0°C~ 45°C (32°F~113°F)
Operating Temperature -20°C ~ 80°C (-4°F~176°F)
Relative Humidity 10%~90% (non-condensing)
Vibration mSATA: 5 grms / 5~500Hz / random operation
Certification CE, FCC class A
Operating System Win10 64-bit Enterprise
SC550N4 SDI Capture card Specifications Specifications
Max. FPS 1920×1080p @60/50fps in → 1920×1080p @60/50fps out
Recording Mode Software Compression, Real-Time Mode
Dimension 130.49 x 101.02 ( mm ) PCI-E Full Height
Interface PCI-E×4 ( Gen2 )
Video Input 4x SDI
Display Video Format YV12, NV12, YUY2, RGB24, RGB32
Video RAW Data Resolution 3G-SDI:
1920 x 1080p @60/50fps
HD-SDI:
1920 x 1080p @30/25/24fps
1920 x 1080i @60/50fps
1280 x 720p @60/50fps
SD-SDI:
720 x 480i @60fps
720 x 576i @50fps
Recording Video Format H.264 ( Software Compression )
Recording Video Resolution 3G-SDI:
1920 x 1080p @60/50fps
HD-SDI:
1920 x 1080p @30/25/24fps
1920 x 1080i @60/50fps
1280 x 720p @60/50fps
SD-SDI:
720 x 480i @60fps
720 x 576i @50fps
Audio Input 4x SDI Embedded Audio
PCB Audio Pin Input: 4x Stereo Audio
Audio Format Stereo / 16-bit / 32000 ~ 48000Hz
I/O 16
WatchDog Yes
SDK Support API : DirectShow, V4L2, FFmpeg, Gstreamer
Support Language : C++, C#, .NET, Visual Basic, Qt, Delphi
OS Support Windows 7 / Windows 8 / Windows 8.1/ Windows 10
Linux 2.6.14 or Higher ( 32-bit and 64-bit )
GEN3G-200 Camera Specifications Specifications
Image Sensor 1/3 inch Progressive CMOS
Sensor Size 5.346 x 3.003mm (16:9)
Effective Pixels Horrizontal: 1920, Vertical: 1080
Pixel Size 2.75um x 2.75um
Resolution 1080p / 1080i: 1920 x 1080
720p & 720i: 1280 x 720
Frame Rate 1080 60p/ 59.94p/ 50p/ 30p/ 29.97p/ 25p
1080 60i/ 59.94i/ 50i
720 60p/ 59.94p/ 50p/ 30p/ 25p
Output HD-SDI/ 3G-SDI (1 BNC)
HDMI1.3a
Color Space 4:2:2 (YCbCr) 10bit
Camera Control RS-485, OSD Button
Dimension 49mm x 49mm x 71mm
1.92” x 1.92” x 2.79”
(With default 3.6mm Lens)
Power 6W at 12 VDC (+9 ~ +15VDC)
SI-62S-E Box PC Controller, w/ Intel® Core™ i7-8700, 4GB x2 DDR4 SO-DIMM, 1x 2.5" 128GB MLC Industrial-grade SSD, 1x SC550N4 SDI capture card, 2x AIDA GEN3G-200 Camera
SI-62S-S Server PC Controller, w/ Intel® Xeon® E-2278GEL, 8GB x2 DDR4, 1x 2.5" 128GB MLC industrial-grade SSD, 1x 2.5" 2TB HDD, 1x SC550N4 SDI capture card
DATASHEETS
SI-62S-E / SI-62S-S
Product Brief
SI-62S-E / SI-62S-S

RELATED PRODUCTS