Edge ML

ML: Machine Learning

Edge machine learning studies how to run useful prediction and learning directly on devices such as sensors, phones, microcontrollers, and embedded accelerators. Common approaches include quantization, pruning, distillation, and hardware-aware architecture search, alongside federated learning and cloud-edge collaboration when devices need to share updates or offload computation. Applications range from health and agriculture to industrial monitoring, with growing work on continual learning and on-device fine-tuning as conditions change. The central challenge is balancing accuracy, energy, memory, latency, privacy, and robustness, while evaluating systems on their actual hardware and under realistic data and network shifts.

Latest papers

130 papers12 papers in the last four weeks, up 20% on the four weeks before. 0.1% of all new papers.

TrainDeeploy: Hardware-Accelerated Parameter-Efficient Fine-Tuning of Small Transformer Models at the Extreme Edge
Mar 10, 2026Run Wang, Victor J. B. Jung, Philip Wiese +3On-Device Training Edge ML
Integrated Systems Laboratory (IIS), ETH Zurich, Switzerland · Department of Electrical, Electronic and Information Engineering (DEI), University of Bologna, Italy · Department of Control and Computer Engineering (DAUIN), Politecnico di Torino, Italy
Two Teachers Better Than One: Hardware-Physics Co-Guided Distributed Scientific Machine Learning
Mar 10, 2026Yuchen Yuan, Junhuan Yang, Hao Wan +4Scientific ML Edge ML
George Mason University, Fairfax, VA, USA · University of North Carolina at Chapel Hill, NC, USA
TT-SEAL: TTD-Aware Selective Encryption for Adversarially-Robust and Low-Latency Edge AI
Feb 24, 2026Kyeongpil Min, Sangmin Jeon, Jae-Jin Lee +1Privacy-Preserving ML Edge ML
†Department of Intelligent Semiconductor Engineering, Chung-Ang University, Seoul, Korea
Analyzing Latency Hiding and Parallelism in an MLIR-based AI Kernel Compiler
Feb 22, 2026Javed Absar, Samarth Narang, Muthu BaskaranHigh-Performance Computing Edge ML
Qualcomm Technologies International, Ltd. · Qualcomm Technologies, Inc.
The Weight of a Bit: EMFI Sensitivity Analysis of Embedded Deep Learning Models
Feb 18, 2026Jakub Breier, Štefan Kučerák, Xiaolu HouNeural Network Robustness Edge ML
TTControl GmbH, Vienna, Austria. · the Faculty of Informatics and Information Technologies, Slovak University of Technology, Slovakia. · the State Key Laboratory of Blockchain and Data Security , Zhejiang University.
U-FedTomAtt: Ultra-lightweight Federated Learning with Attention for Tomato Disease Recognition
Feb 18, 2026Romiyal George, Sathiyamohan Nishankar, Selvarajah Thuseethan +4Lightweight Neural Networks Federated Learning
Department of Computer Engineering, University of Peradeniya, Sri Lanka. · Faculty of Science and Technology, Charles Darwin University, Australia. · Canadian Institute for Cybersecurity, University of New Brunswick, Canada. +1
Self-Evolving Multi-Agent Network for Industrial IoT Predictive Maintenance
Feb 17, 2026Rebin Saleh, Khanh Pham Dinh, Balázs Villányi +1Predictive Maintenance Industrial Anomaly Detection
Department of Electronics Technology, Faculty of Electrical Engineering and Informatics, Budapest University of Technology and Economics, Műegyetem rkp. 3, H-1111 Budapest, Hungary · DataScienceWorld.Kan, Hanoi, Vietnam · Department of Computer Science, The University of Alabama at Birmingham, 1720 University Blvd, Birmingham, AL 35294 USA
Triplet Feature Fusion for Equipment Anomaly Prediction : An Open-Source Methodology Using Small Foundation Models
Feb 16, 2026Takato YasunoAnomaly Detection Predictive Maintenance
AutoTailor: Automatic and Efficient Adaptive Model Deployment for Diverse Edge Devices
Nov 27, 2025Mengyang Liu, Chenyu Lu, Haodong Tian +7Edge ML Efficient Neural Network Inference
Southeast University · Hong Kong University of Science and Technology · Tsinghua University +1
BiHDTrans: binary hyperdimensional transformer for efficient multivariate time series classification
Sep 29, 2025Jingtao Zhang, Yi Liu, Qi Shen +1Efficient Neural Network Inference Transformer
School of Biomedical Engineering, Shenzhen Campus of Sun Yat-sen University · College of Medicine and Biological Information Engineering, Northeastern University

Mar 10, 2026J/K move · Enter open · S save

TrainDeeploy: Hardware-Accelerated Parameter-Efficient Fine-Tuning of Small Transformer Models at the Extreme Edge

Run Wang, Victor J. B. Jung, Philip Wiese, Francesco Conti+2

Integrated Systems Laboratory (IIS), ETH Zurich, Switzerland · Department of Electrical, Electronic and Information Engineering (DEI), University of Bologna, Italy · Department of Control and Computer Engineering (DAUIN), Politecnico di Torino, Italy

On-Device Training Edge ML Memory-Efficient Fine-Tuning Parameter-Efficient Fine-Tuning

Open paper AskPDF arXiv

On-device tuning of deep neural networks enables long-term adaptation at the edge while preserving data privacy. However, the high computational and memory demands of backpropagation pose significant challenges for ultra-low-power, memory-constrained extreme-edge devices. These challenges are further amplified for attention-based models due to their architectural complexity and computational scale. We present TrainDeeploy, a framework that unifies efficient inference and on-device training on heterogeneous ultra-low-power System-on-Chips (SoCs). TrainDeeploy provides the first complete on-device training pipeline for extreme-edge SoCs supporting both Convolutional Neural Networks (CNNs) and Transformer models, together with multiple training strategies such as selective layer-wise fine-tuning and Low-Rank Adaptation (LoRA). On a RISC-V-based heterogeneous SoC, we demonstrate the first end-to-end on-device fine-tuning of a Compact Convolutional Transformer (CCT), achieving up to 11 trained images per second. We show that LoRA reduces dynamic memory usage by 23%, decreases the number of trainable parameters and gradients by 15x, and reduces memory transfer volume by 1.6x compared to full backpropagation. TrainDeeploy achieves up to 4.6 FLOP/cycle on CCT (0.28M parameters, 71-126M FLOPs) and up to 13.4 FLOP/cycle on Deep-AE (0.27M parameters, 0.8M FLOPs), while expanding the scope of prior frameworks to support both CNN and Transformer models with parameter-efficient tuning on extreme-edge platforms.