diff --git a/.wordlist.txt b/.wordlist.txt index 3d9a0888..00a8f2fb 100644 --- a/.wordlist.txt +++ b/.wordlist.txt @@ -1784,4 +1784,5 @@ SonicSight FedAvg EfficientNetB generalizable +wakeword diff --git a/README.md b/README.md index 501893a1..78587786 100644 --- a/README.md +++ b/README.md @@ -103,7 +103,7 @@ Audio classification, keyword spotting, wakeword detection, or other machine lea ### Predictive Maintenance & Fault Classification -Projects devoted to the use of sensors, audio, or image data specfic to the predictive maintenance use-case. +Projects devoted to the use of sensors, audio, or image data specific to the predictive maintenance use-case. * [Predictive Maintenance - Nordic Thingy:91](predictive-maintenance-and-fault-classification/predictive-maintenance-with-nordic-thingy91.md) * [Brushless DC Motor Anomaly Detection](predictive-maintenance-and-fault-classification/brushless-dc-motor-anomaly-detection.md) @@ -173,3 +173,4 @@ Using machine learning to predict, understand, or identify information in the ai * [Using Hugging Face Image Classification Datasets with Edge Impulse](software-integration-demos/hugging-face-image-classification-dataset.md) * [Edge Impulse API Usage Sample Application - Jetson Nano Trainer](software-integration-demos/api-sample-application-jetson-nano.md) * [MLOps with Edge Impulse and Azure IoT Edge](software-integration-demos/mlops-azure-iot-edge.md) +* [A Federated Approach to Train and Deploy Machine Learning Models](federated-learning.md) diff --git a/SUMMARY.md b/SUMMARY.md index 3dbc7935..20fa7e74 100644 --- a/SUMMARY.md +++ b/SUMMARY.md @@ -157,3 +157,4 @@ * [Using Hugging Face Image Classification Datasets with Edge Impulse](software-integration-demos/hugging-face-image-classification-dataset.md) * [Edge Impulse API Usage Sample Application - Jetson Nano Trainer](software-integration-demos/api-sample-application-jetson-nano.md) * [MLOps with Edge Impulse and Azure IoT Edge](software-integration-demos/mlops-azure-iot-edge.md) +* [A Federated Approach to Train and Deploy Machine Learning Models](federated-learning.md) diff --git a/software-integration-demos/federated-learning.md b/software-integration-demos/federated-learning.md index 45a35dff..79e9b64b 100644 --- a/software-integration-demos/federated-learning.md +++ b/software-integration-demos/federated-learning.md @@ -175,7 +175,7 @@ Finally, when all the images have been labeled, we can click "Model testing" and ## Result -Finally, after training a decentralized model and uploading it to Edge Impulse, one incredible feature that we can benefit from is a seamless deployment of the model on hardwares ranging from MCUs, CPUs and custom AI accelerators. In this case, we can deploy our model to the Raspberry Pi as an [.eim executable](https://docs.edgeimpulse.com/docs/tools/edge-impulse-for-linux#.eim-models) that contains the signal processing and ML code, compiled with optimizations for a processor or GPU (e.g. NEON instructions on ARM cores) plus a very simple IPC layer (over a Unix socket). +Finally, after training a decentralized model and uploading it to Edge Impulse, one incredible feature that we can benefit from is a seamless deployment of the model on hardware ranging from MCUs, CPUs and custom AI accelerators. In this case, we can deploy our model to the Raspberry Pi as an [.eim executable](https://docs.edgeimpulse.com/docs/tools/edge-impulse-for-linux#.eim-models) that contains the signal processing and ML code, compiled with optimizations for a processor or GPU (e.g. NEON instructions on ARM cores) plus a very simple IPC layer (over a Unix socket). First, we need to attach the Raspberry Pi camera to the to the board.