Solution

Artificial Intelligence SolutionDeep-Learning & Edge Solution

Deep-Learning Solution in small-size, low-power module with FPGA without GPU Server & Software

Developing deep-learning visual sensor
Deep-Learning Solution
Object detect & categorize
Object detect & categorize
Object detect & categorize
Recognition range
Object detect & categorize Person/Gender/Age/Emotion
Benefit
  • Small, low-power deep learning module with FPGA
  • Detect & categorize About 190 objects
  • Reinforcement learning using Deep Learning
Product category
  • 4x5cm size
  • unit setting in user
  • Video input signal: YCbCr4:2:2
  • Output: ethernet
  • Power: 5W
  • Video I/O: HDMI
  • Output: Ethernet, pin out
  • Split Screen: 16 channels
  • Power: 8W
  • CCTV unmanned only
  • Input: ethernet (IP camera)
  • 8/16Camera real-time recognition
  • NVR Recording
SSD300/MobileNet - object recognition algorithm device