Tinyml
PulseAugur coverage of Tinyml — every cluster mentioning Tinyml across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
New multi-exit TinyML scheme boosts edge AI efficiency
Researchers have developed a novel multi-exit computational scheme for TinyML systems on edge devices, aiming to improve energy efficiency and real-time performance. This approach, deployed on a GWT GAP9 System-on-Chip,…
-
INT8 Quantization Shrinks TinyML ECG Model by 60%
An experiment explored the impact of INT8 quantization on a TinyML model designed for ECG arrhythmia detection. By reducing the numerical precision from 32-bit floating point (FP32) to 8-bit integers (INT8), the model s…
-
TinyML Systems Explore Instance Hardness for Energy Efficiency
Researchers have presented preliminary findings on a new application of the tree depth prune instance hardness method within TinyML systems. This approach aims to reduce computational costs and energy consumption for AI…
-
AI models for text recognition reviewed: challenges and future directions
A recent literature review, adhering to PRISMA guidelines, analyzes 97 studies from January 2015 to January 2025 on machine learning models for optical character recognition (OCR). The review details the evolution of AI…
-
TinyML system PolyChirp enables multi-species bird classification
Researchers have developed PolyChirp, a novel approach for classifying multiple bird species using TinyML on low-power acoustic sensors. This system is designed to overcome the limitations of previous TinyML models, whi…
-
TinyML Systems Performance and Power Characterization Detailed
This paper provides a detailed analysis of the performance and power consumption of TinyML systems deployed on microcontrollers. It investigates the trade-offs between programmability and efficiency across various abstr…
-
New framework uses LLMs to select explainable AI for TinyML edge devices
Researchers have developed a new framework for selecting explainable AI (XAI) methods for TinyML edge devices, particularly for clinical applications. This framework uses a large language model (LLM) to guide the design…
-
Bio-Tuning Glasses: Invisible Biofeedback Interface Adapts Environment Using Edge AI
Researchers are developing Bio-Tuning Glasses, an experimental concept for an invisible biofeedback interface that adapts the user's environment based on their physiological state. Unlike typical wearables that notify u…
-
Tiny robot navigation enhanced with compact Nano-U terrain segmentation model
Researchers have developed Nano-U, a compact neural network designed for efficient terrain segmentation on low-cost microcontrollers. This approach addresses the limitations of current models for small robotic platforms…
-
ColabNAS offers affordable HW NAS for lightweight CNNs
Researchers have developed ColabNAS, an accessible hardware-aware neural architecture search (HW NAS) technique designed to create lightweight, task-specific convolutional neural networks (CNNs). This method, inspired b…
-
New HDC Framework Enhances Anomaly Detection for Edge AI
Researchers have introduced D2H-AD, a novel anomaly detection framework that leverages Hyperdimensional Computing (HDC). This brain-inspired approach uses high-dimensional vectors to represent information, integrating d…
-
New EFGCN processes event data on FPGAs with 100x smaller models
Researchers have developed an embedded graph convolutional network (EFGCN) specifically designed for real-time event data processing on System-on-Chip (SoC) FPGAs. This approach significantly reduces model size, by up t…
-
TinyML models analyzed for spacecraft cybersecurity
A new research paper analyzes the performance of TinyML models for cybersecurity threats on autonomous spacecraft. The study focuses on the latency-accuracy trade-offs of classical machine learning models like Random Fo…
-
In-sensor computing boosts satellite Earth observation efficiency
Researchers have developed a new in-sensor computing framework for energy-efficient Earth observation from satellites. This approach integrates TinyML techniques with the Sony IMX500 Intelligent Vision Sensor to process…
-
TinyML models enable on-device arrhythmia detection
Researchers have developed ArrythML, a TinyML approach for on-device arrhythmia detection using autoencoder models. These INT8 quantized models are designed for resource-constrained embedded systems, processing over 95,…
-
TinyML survey highlights on-device learning challenges
A new survey paper published on arXiv examines the challenges of on-device learning (ODL) for TinyML applications. It highlights how changes in data distribution after deployment can degrade the performance of static mo…
-
Ariel-ML toolkit enables Rust-based parallel neural network inference on multi-core microcontrollers
A new toolkit named Ariel-ML has been developed to automate parallelization for neural network inference on multi-core microcontrollers using embedded Rust. This toolkit is designed to leverage the capabilities of heter…
-
AI framework boosts energy efficiency for smart city environmental monitoring
Researchers have developed an AI-driven framework designed to make environmental monitoring in smart cities more energy-efficient. This system utilizes TinyML-enabled edge devices that dynamically activate sensors based…
-
Researchers develop browser-based and on-device TinyML vision training
Two research papers detail novel approaches for training and deploying machine learning vision models directly on low-cost microcontrollers. One paper introduces a browser-based application that facilitates a complete, …