AI Products / Services > Professinal Service

AI Solution

header image

Overview

DMP provides optimal AI solutions independent of platform and specialized for specific domains.

For the tasks to be solved related to AI and computer vision, DMP proposes optimal solutions among AI technologies that evolve day by day, based on academic papers or using original technology. DMP implements optimal software for various devices such as FPGA, CPU (Intel, ARM, etc.), GPU (NVIDIA, AMD), and SoC (Qualcomm Snapdragon), realizes high speed and provides the optimal platform according to the customer's requirements.
DMP provides various services such as annotation service, generation of synthetic data set, neural network design/training, driver development/device implementation, system integration/application development, and system evaluation/performance evaluation.

Specific Implementing Function Examples

Privacy protection of medical diagnostic images

Autonomous driving system for low speed vehicles

Implementation Steps

DMP provides various services such as parts selection, circuit design, board layout, simulation, and system verification.

Customer's product development process
1. Survey and Planning
  • Propose the optimal AI utilization plan for the customer's request
  • Propose solutions to tasks from both software and hardware sides for PoC development
2. PoC Development
  • Collect actual data from customers, or collect and create necessary data itself, performs annotations, and create learning model
  • Evaluate the performance and recognition accuracy of the created learning model
    Achieve early PoC development by utilizing proven trained models
  • CPU, GPU, and FPGA can be adopted for the hardware platform for PoC development
3. Product Development
  • Design learning model optimal for customer's requirements such as performance, recognition accuracy and implement it in the targeted device
  • Realize further speedup and power saving through logic design of SoC/FPGA for hardware acceleration in case customers develop SoC/FPGA
4. Operation after Product Implementation
  • Regular re-training and optimization possible
  • It is possible to update the trained model according to changes in customer's requirements on recognition accuracy and performance.
expand_less