Edge Impulse

Additional features

AI for Any Edge Device

Enables AI development for MCUs, NPUs, CPUs, GPUs, gateways, sensors, cameras, and Docker containers.

Dataset Building

Allows users to create and manage datasets.

Model Training

Provides capabilities to train machine learning models.

Library Optimization

Optimizes libraries to run directly on devices.

Intelligence on Edge Devices

Brings AI capabilities to edge devices.

Sensor Data Unlocking

Helps in extracting insights from sensor data.

Faster Time to Market

Accelerates the delivery of next-generation products and solutions.

Complexity Removal

Simplifies AI development by removing hidden complexities and repetitive steps.

High-Value Task Progression

Enables quick progress on tasks leading to commercialization.

De-risked Model Development

Offers agnostic and scalable edge AI tools to reduce development risks.

Manufacturing and Operations Edge AI

Addresses issues like downtime, inefficiency, and quality control through early anomaly detection.

Product Development Edge AI

Fast-tracks development and increases successful edge AI deployment.

Transportation Edge AI

Provides solutions for smart city intersections, vehicle safety, railway safety, and more.

Industrial Edge AI

Adds new insights to sensor networks through embedded machine learning for industrial enterprises.

Any Data Support

Capable of processing various data types (time-series, audio, image data).

Any Model Support

Supports various model types and training techniques.

Seamless Collaboration

Facilitates collaboration for production-ready models.

Web-based Studio

Offers a graphical interface for data collection, impulse building, and deployment.

Data Acquisition Tool

Manages storing, sorting, and labeling of data.

Impulse Design

Allows creation of impulses with feature extraction methods and ML models.

Pre-built Feature Extraction Methods

Provides off-the-shelf methods for common sensor data (e.g., Raw Data, Flatten, Image, Spectrogram, Audio MFE/MFCC, IMU, HR/HRV features).

Custom Processing Blocks

Enables users to design their own feature extraction methods.

Learning Blocks

Supports training of classification, regression, or anomaly detection models.

Custom Learning Blocks

Allows creation of custom learning blocks.

Expert Mode

Provides options to modify ML training code.

Model Testing

Supports testing models using a holdout set or live device data.

Deployment Options

Deploys full impulses in various formats (C++ library, Linux process, Docker container, WebAssembly, pre-built firmware).

EON Tuner (AutoML)

Automatically tries various impulse configurations to determine the best combination for on-device performance.

Web API Scripting

All aspects of the Studio can be scripted using a web API for MLOps pipelines.

Organization Hub

For enterprise clients, easily monitor and maintain projects.

Automated Data Pipelines

Configure and run transformation blocks to extract, transform, and load (ETL) data.

Feature Importance Analysis

Helps determine which features are most significant for analysis.

Data Explorer

Visualizes raw dataset data.

Feature Explorer

Visualizes features generated by processing blocks to identify outliers and class separation.

Anomaly Detection Algorithms

State-of-the-art algorithms for detecting anomalies in all sensor modalities.

Object Detection (YOLO-Pro)

Specialized model for industrial object detection.

Application Behavior Analysis

Understands and improves model performance in real-world scenarios.

Model Monitoring

Continuously enhances models with real-world data in production.

Synthetic Data Generation

Tools to generate synthetic data for various data types.

Hardware Profiling Tools

Assess memory, flash, and latency requirements for specific devices.

DSP and Transformation Compilation

Compiles code specifically for target devices.

Hyperparameter Tuning

Tools for optimizing model parameters for specific devices.