We reduce the time and effort required for model improvement with our unique neural architecture search technology.
■Do you have any of the following challenges? - The processing time of the DNN model is long, making real-time processing difficult. - I want to lightweight and speed up the model for implementation on edge devices. - I want to avoid a decrease in image recognition accuracy due to acceleration. - A lot of effort is required for model modification, measurement, and evaluation. - I feel limited in manually searching for the optimal DNN model. - I want to shorten the AI development period and accelerate the market launch of products and services. - I cannot find a model that balances the desired recognition accuracy and processing performance. Our DNN acceleration service supports the acceleration of image-dependent processing while maintaining the accuracy of DNN models. ■Features of the DNN Acceleration Service ✓ Accelerates DNN models while maintaining recognition accuracy. ✓ Automatically generates and selects numerous DNN models. ✓ Reduces the development time and effort required for model improvement.
Inquire About This Product
basic information
■What is the DNN Acceleration Service? Our unique automation technology accelerates the image recognition DNN models entrusted to us without compromising accuracy, all within a short period. < Point 1 > Balancing Speed and Accuracy We automatically create numerous modified models that enhance processing performance with minimal impact on inference accuracy from the DNN models provided by our customers. By conducting evaluations for each modified model, we deliver the most suitable DNN model that meets the target processing performance and inference accuracy. < Point 2 > Reducing Workload with Automatic Model Generation Technology For the automatically generated DNN models, we perform performance and accuracy predictions to automatically filter out modified models that are expected to fall short of the targets. Only the selected modified models undergo performance measurement and inference accuracy evaluation, allowing us to explore optimal models in a short timeframe. ■Provision Format We provide DNN models that meet the target accuracy and performance by receiving DNN models, training data, and accuracy evaluation metrics.
Price information
Please feel free to contact us with any questions or for a quote. Our products and services are sold to companies, organizations, and corporations.
Delivery Time
Applications/Examples of results
■Application Example Reduced processing time of the semantic segmentation model by approximately 40% This is an example of applying this service to U-Net, a representative network for semantic segmentation. - Processing time: reduced from 37.81ms to 22.31ms - Processing time: approximately 40% reduction - Accuracy metric mIoU: increased from 58.93% to 59.60% - Reduced processing time while maintaining accuracy As an example of accelerating inference processing while maintaining image recognition accuracy, this serves as a reference when considering implementation in image recognition systems that require real-time performance or in devices with processing performance constraints. The final description of the application is a suggested expression to make the example easier to understand.
catalog（1）
Download All Catalogs
Company information
We will provide optimal solutions that contribute to our customers' businesses through our "extensive experience and achievements accumulated over many years" and "high technical capabilities" in the fields of embedded and LSI design.
![[IGOR Pro Case Study] University of Tokyo, Atmosphere and Ocean Research Institute](https://image.mono.ipros.com/public/product/image/0e2/2001491923/IPROS64344024876720016148.png?w=280&h=280)
