Scaleout Edge: Sovereign Edge AI orchestration and Federated Learning
In this repository, we explore model compression for transformer architectures via quantization. We specifically explore quantization aware training of the linear layers and demonstrate the performance for 8 bits, 4 bits, 2 bits and 1 bit (binary) quantization.
A camera for measuring sediment grain sizes with edge ML
ESP32-S3 WiFi CSI Human Activity Recognition system with real-time RF sensing, ML inference, and live visualization dashboard.
Edge-computed multi-sensor platform correlating hydrophone acoustics with environmental data (temp/pH/turbidity/salinity) for marine ecosystem health monitoring.
A curated list of small foundation models (SFMs) designed to run locally on Edge devices.
BEAVER automates the Edge AI lifecycle through LLM-powered orchestration. It Builds, Evolves, Analyzes, Validates, Executes, and Repairs—just like beavers in nature, but for edge devices!
Edge-deployable keyword spotter: INT8-quantized DS-CNN on Google Speech Commands, exported to ONNX, with fp32 vs INT8 benchmarks, a live mic demo, and a C++ inference harness.
Notes and resources from Qualcomm On-device AI course, provided by DeepLearningAI
Compress PyTorch models for edge devices — CPU-only, no GPU, no retraining. One function call.
embedded software components for event-based application development
Python ML library for person fall detection. Intended for IoT deployments with on-device inference and on-device transfer learning.
Lightweight Attention U-Net for Breast Cancer Semantic Segmentation
Fault-tolerant Edge ML pipeline for space weather (TEC) prediction. Fuses raw ISRO telemetry with NASA APIs for sub-second, on-device inference via quantized TensorFlow Lite.
Edge-first XRF V2 benchmark for deployable wearable event detection (earbuds + smart-glasses), with FP/hour-calibrated metrics and reproducible run artifacts.
A system for monitoring statistical data distribution shifts in distributed settings
Pre-quantized models for edge inference on Cortex-M, ESP32-S3, and Raspberry Pi. Six models — keyword spotting, person detection, hand-gesture recognition, anomaly detection, voice activity detection, and binary defect classifier — each shipped as TF
A curated list of machine learning models that run locally on edge devices, including phones, browsers, laptops, Raspberry Pi-class boards, Jetson, Coral, NPUs, and microcontrollers.
Benchmarking the accuracy, size, latency, and memory trade-offs of post-training quantization across four model families and three modalities, with one reproducible harness that runs the same on x86 and ARM64. TensorFlow and PyTorch, TFLite and ONNX Runtime.
Add a description, image, and links to the edge-ml topic page so that developers can more easily learn about it.
To associate your repository with the edge-ml topic, visit your repo's landing page and select "manage topics."