Overview
Product Information on Edge Impulse
What is Edge Impulse?
Edge Impulse Pricing
Overall experience with Edge Impulse
“Product Praised for Ease but Improvement Needed in IAR Library Integration”
“Excellent tool with some issues in deployment to big embedded systems”
About Company
Company Description
Edge Impulse is a software platform that helps companies run machine learning models on real world devices that make decisions and offer insights where data is gathered. The platform accelerates time to market, improves ML outcomes, and de-risks on-device deployment so users can develop and productize next generation products and solutions. The platform offers the power to seamlessly manage machine learning for the edge, including data acquisition, signal processing, ML training, model testing, and creating a deployable model that can run efficiently on an edge device from MCU to CPU to GPU.
Company Details
Do You Manage Peer Insights at Edge Impulse?
Access Vendor Portal to update and manage your profile.
Key Insights
A Snapshot of What Matters - Based on Validated User Reviews
Reviewer Insights for: Edge Impulse
Performance of Edge Impulse Across Market Features
Edge Impulse Likes & Dislikes
1) Ease of use 2) Support and resources 3) Useful features
Ease of use, good documentation and great API
It's versatile, hardware agnostic, user-friendly
- Lack of layer normalization - Lack of learning rate scheduler - Generic IAR library needs some work to get going
The customized libraries from other sources cause headache. Symbol duplication in causing issues in the end product as TensorFlowMicro uses same functions, but they get added many times and then the ROM consumption increases.
Overall it's quite intuitive, but some functions need unnecessarily many clicks. From time to time training is much slower than average. It is well documented, but the documentation is a bit scattered.
Top Edge Impulse Alternatives
Peer Discussions
Edge Impulse Reviews and Ratings
- Engineer1B-10B USDManufacturingReview Source
Product Praised for Ease but Improvement Needed in IAR Library Integration
Helpful team. Quick support. Able to develop and iterate very efficiently - Senior LecturerGov't/PS/EdGovernmentReview Source
Versatile Edge AI Platform Offers Hardware Agnostic Functionality Across Devices
The product is very versatile, you can do pretty much anything related to Edge AI, regardless of device. - SW Architect50M-1B USDConsumer GoodsReview Source
Excellent tool with some issues in deployment to big embedded systems
The service works super great in data handling and training the models and the performance of the output is great. Some issues are with warnings from the delivery object and some difficulties with multi-impulse feature. - Staff Engineer10B+ USDEnergy and UtilitiesReview Source
Prohibitive Pricing for PoCs, But Excellent Ease of Use and Interface
I had a wonderful experience utilizing the platform to build AI models for the company. There were deficiencies back when I used the platform. It had no support for classical models. I hope it has changed now. - Director of IT<50M USDIT ServicesReview Source
Ease of use and deployment options to edge
User friendly interface with good documentation. No issues encountered so far