Edge Impulse Reviews

I enjoys how easy it is to import and scale data. Overall Edge Impulse is a platform. It could be simplified for everyday users and offer more options for diverse applications of knowledge.

Whenever you retrain the model there are changes in the calculated accuracy results. This makes it difficult to rely on Edge Impulses classification accuracy consistently.

Their platform simplifies machine learning model training and deployment onto our devices significantly enhancing accuracy and performance in our systems. As a result our business benefits, from informed decision making. An improved overall user experience.

The platform provides a range of data augmentation and preprocessing tools that help enhance the quality of your training data. These features can be particularly beneficial when dealing with limited datasets, as they contribute to better model performance and generalization.

Although Edge Impulse provides a wealth of documentation and resources, some users might find it helpful to have offline documentation available for situations when they're working in environments with limited internet access.

Developing edge AI applications involves iteration and experimentation. Edge Impulse accelerates prototyping by offering a seamless workflow, allowing developers to quickly build, train, and test models on real-world data.

The platform offers various deployment options, including exporting models to different formats that are optimized for specific edge devices. This flexibility ensures that your models can be efficiently utilized across your target hardware.

The platform might feel somewhat limited in terms of building very complex or specialized models. Users with advanced machine learning needs might desire more extensive customization options or support for more advanced model architectures.

Edge Impulse streamlines the complex process of developing machine learning models for edge devices. It provides an integrated platform that covers data collection, preprocessing, model training, deployment, and monitoring. This simplification saves developers significant time and effort compared to manually piecing together different tools and frameworks.

appreciate Edge Impulse for its user-friendly interface, machine learning capabilities tailored for edge devices, and its support for developing and deploying AI models for various applications

There are some limitations in terms of certain advanced customization options, compatibility with specific hardware, or the learning curve for those who are new to machine learning and IoT technologies.

Data Collection and Labeling: Edge Impulse simplifies the process of collecting and labeling sensor data, which is crucial for training accurate models. This benefits me by saving time and effort in manually managing and preparing my data.

Edge impulse makes machine learning possible for Edge devices using an easy Graphical User Based interface. It supports a wide range of devices such as Rasberry Pie, Mobile phones etc.

Support for custom embedded devices could be enhanced. The company could provide a few more features in the developer Edition so that more and more people can try it.

Used this solution for a small project on Embedded devices. The development was quick as it provided a lot of GUI components for drag and drop. We had to write no code.

Real time problems can be easily solved with the help of machine embedded system.

Machine learning is not everyone's cup of tea.

With the help of embedded sysytem MLOps are sloved within 5 minutes of traction time.

The user interface which lets a new user to interact and use the platform with confidence without hiding anything from user is one of the best things I like about the Edge Impulse

The interface on the mobile is needs to get improved compared to desktop view. Also the solution section could have more options to access variety and user can apply the knowledge with diversification.

The Edge Impulse is specifically used for Machine Learning which enables the deployment on various hardware to get amazing results and next to accurate predictions. And this benifts me to understand the ML deployment on various hardware.

The look and feel and the ease of use. I like the simple and logical way the workflow is organised.

Minimal data import and data scaling abilities. One big problem I had was the time scaling, which is restricted to milliseconds - this is not helpful when importing data from many time series databases

AI has great potential for improving the accuracy of air quality sensor measurement data. One use case would be auto-calibration of low cost sensors who high accuracy sensors

It is a Machine learning based platform for businesses to enhance their experience on different embedded devices for audio-video visuals, sensors on a large scale. Through it's help, all the engineers or developers can solve the problems using Machine learning leading to the solution time to be very quick.

The platform in itself is good enough to use but it can be simplified and made easy to understood by common people also in terms of general analytics rather than just the developers or coders

Edge impulse is machine learning based platform, so giving the most optimum and accurate outputs in the minimum time is one of the major outcomes besides reducing the human efforts

Edge impulse is one of the emerging platforms for embedded ML and is free for developers. Without specialised hardware like Arduino or raspberry pi, many real-time and time-consuming machine learning problems can be solved quickly. The best thing about this is you can use your mobile phone, computer or a supported development board. The data collection is super easy, and the dashboards help you manage your data. For example- you can view your previous data, projects and devices you used to connect etc. Very easy to use, just scan the QR code and the device gets connected via a link. Like any ML model, train and test data sets need to be created and you can create various labels for your reference. You can add filters to improve your results or for the desired conversion or input. It's straightforward; even if you are a beginner, just add the recommended filters, and they work great. Drag and drop approach solves most of the coding approach. There aren't many open source GUI-based approaches such as this one.

Not much really. It reduces the coding approach, which might be a habit for developers over time. But everyone is looking for GUI models these days, so might be helpful. On the other hand, it can be difficult for anyone looking for control of every single parameter and full customisation to the details of the approach.

For me, it was the accuracy issue. Every time you retrain your model, there are some internal changes to the previously calculated accuracy results by Edge Impulse, based on that, it creates another good classifier.