TI OMAP4, Qualcomm Snapdragon MSM8960 & MSM8974 and Redhook Board
Providing Low-Power FPGA Design Solutions for Deep Learning
The market for low cost, high-performance FPGA design solutions is growing increasingly competitive. As a new generation of Edge applications emerges, designers are increasingly pressed to develop solutions that combine low power and small form factor and still achieve increased performance demands.
The leading U.S. based low-cost programmable logic provider offering solutions across the network from the Edge to the Cloud across the industries such as growing communications, computing, industrial, automotive and consumer markets was looking to build deep learning solutions on its platform.
Softnautics took ownership of end-to-end solution reference design validation with dataset identification, tuning, and cleanup as per the client's requirement. The technology expertise and model tuning experience by Softnautics enabled the client to unleash the potential of its platform to showcase a real-time smart vehicle classification system.
The client was looking for a system which can provide a value in terms of vehicle monitoring and identification based on vehicle appearance in addition to the vehicles’ attached license plate typical recognition.But challenged faced with the train models which were compatible with custom Convolution Neural Network IP implementation, RTL dependencies for end-to-end use cases on lower power FPGA.
Low-cost ultra-low power configurable solution to support classification categories
End to end solution stack is provided enabling fast time-to-market
Trained model to classify vehicles as per height categories
Real-time object detection and height estimation
The solution can be easily adapted to
• Face applications such as detection, recognition, tracking, tagging, etc.
• Analog Gauge Meter reader for industrial automation
• Defect detection in manufacturing
• Image analytics for security & surveillance
• Voice Assistants
Tools and Technologies
SLIMBUS Analyzer, CRO