mlsanityData to 14 Oct 2026

Enabling Design Methodologies and Future Trends for Edge AI: Specialization and Co-design

arXiv 2103.15750 ↗First posted 25 Mar 2021cs.AR cs.AIcited by 13

Not on the map. Why?

Abstract (arXiv)

Artificial intelligence (AI) technologies have dramatically advanced in recent years, resulting in revolutionary changes in people's lives. Empowered by edge computing, AI workloads are migrating from centralized cloud architectures to distributed edge systems, introducing a new paradigm called edge AI. While edge AI has the promise of bringing significant increases in autonomy and intelligence into everyday lives through common edge devices, it also raises new challenges, especially for the development of its algorithms and the deployment of its services, which call for novel design methodologies catered to these unique challenges. In this paper, we provide a comprehensive survey of the latest enabling design methodologies that span the entire edge AI development stack. We suggest that the key methodologies for effective edge AI development are single-layer specialization and cross-layer co-design. We discuss representative methodologies in each category in detail, including on-device training methods, specialized software design, dedicated hardware design, benchmarking and design automation, software/hardware co-design, software/compiler co-design, and compiler/hardware co-design. Moreover, we attempt to reveal hidden cross-layer design opportunities that can further boost the solution quality of future edge AI and provide insights into future directions and emerging areas that require increased research focus.

Key facts. Cited by 13 papers in the map (data through 14 Oct 2026). Builds on 65 of its 131 references that are in the map. First posted on arXiv 25 Mar 2021.

Builds on

65 of 131 references.

References · 65

  1. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
  2. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,795
  3. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,845
  4. 2019Searching for MobileNetV3cited by 1,994
  5. 2019Energy and Policy Considerations for Deep Learning in NLPcited by 1,254
  6. 2017ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devicescited by 1,200
  7. 2018ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Designcited by 1,136
  8. 2016SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model sizecited by 1,060
  9. 2018MnasNet: Platform-Aware Neural Architecture Search for Mobilecited by 943
  10. 2019Towards Federated Learning at Scale: System Designcited by 832
  11. 2016EEGNet: A Compact Convolutional Network for EEG-based Brain-Computer Interfacescited by 628
  12. 2019Federated Machine Learning: Concept and Applicationscited by 412
38 more
  1. 2018Slimmable Neural Networkscited by 305
  2. 2020MCUNet: Tiny Deep Learning on IoT Devicescited by 273
  3. 2014cuDNN: Efficient Primitives for Deep Learningcited by 245
  4. 2019Edge Intelligence: Paving the Last Mile of Artificial Intelligence with Edge Computingcited by 228
  5. 2019Improving Neural Network Quantization without Retraining using Outlier Channel Splittingcited by 151
  6. 2018NetAdapt: Platform-Aware Neural Network Adaptation for Mobile Applicationscited by 147
  7. 2020HAT: Hardware-Aware Transformers for Efficient Natural Language Processingcited by 136
  8. 2019Convergence of Edge Computing and Deep Learning: A Comprehensive Surveycited by 120
  9. 2017Bit Fusion: Bit-Level Dynamically Composable Architecture for Accelerating Deep Neural Networkscited by 109
  10. 2018Channel Gating Neural Networkscited by 80
  11. 2020PatDNN: Achieving Real-Time DNN Execution on Mobile Devices with Pattern-based Weight Pruningcited by 76
  12. 2015Conditional Deep Learning for Energy-Efficient and Enhanced Pattern Recognitioncited by 67
  13. 2020EEG-TCNet: An Accurate Temporal Convolutional Network for Embedded Motor-Imagery Brain-Machine Interfacescited by 63
  14. 2018SqueezeNext: Hardware-Aware Neural Network Designcited by 61
  15. 2016ESE: Efficient Speech Recognition Engine with Sparse LSTM on FPGAcited by 52
  16. 2019PCONV: The Missing but Desirable Sparsity in DNN Weight Pruning for Real-time Execution on Mobile Devicescited by 50
  17. 2019FPGA/DNN Co-Design: An Efficient Design Methodology for IoT Intelligence on the Edgecited by 48
  18. 2020Co-Exploration of Neural Architectures and Heterogeneous ASIC Accelerator Designs Targeting Multiple Taskscited by 43
  19. 2019Accuracy vs. Efficiency: Achieving Both through FPGA-Implementation Aware Neural Architecture Searchcited by 38
  20. 2019E2-Train: Training State-of-the-art CNNs with Over 80% Energy Savingscited by 36
  21. 2020EDD: Efficient Differentiable DNN Architecture and Implementation Co-search for Embedded AI Solutionscited by 36
  22. 2019PULP-NN: Accelerating Quantized Neural Networks on Parallel Ultra-Low-Power RISC-V Processorscited by 35
  23. 2019Memory-Driven Mixed Low Precision Quantization For Enabling Deep Network Inference On Microcontrollerscited by 33
  24. 2019Non-Structured DNN Weight Pruning -- Is It Beneficial in Any Platform?cited by 32
  25. 2020FracBNN: Accurate and FPGA-Efficient Binary Neural Networks with Fractional Activationscited by 32
  26. 2019AdaScale: Towards Real-time Video Object Detection Using Adaptive Scalingcited by 27
  27. 2019SkyNet: a Hardware-Efficient Method for Object Detection and Tracking on Embedded Systemscited by 27
  28. 2020AutoDNNchip: An Automated DNN Chip Predictor and Builder for Both FPGAs and ASICscited by 26
  29. 2020A Co-Design Framework of Neural Networks and Quantum Circuits Towards Quantum Advantagecited by 23
  30. 2020FracTrain: Fractionally Squeezing Bit Savings Both Temporally and Spatially for Efficient DNN Trainingcited by 23
  31. 2018Designing Adaptive Neural Networks for Energy-Constrained Image Classificationcited by 18
  32. 2019FANN-on-MCU: An Open-Source Toolkit for Energy-Efficient Neural Network Inference at the Edge of the Internet of Thingscited by 18
  33. 2019Butterfly Transform: An Efficient FFT Based Neural Architecture Designcited by 16
  34. 2020Anti-Bandit Neural Architecture Search for Model Defensecited by 16
  35. 2020VecQ: Minimal Loss DNN Model Compression With Vectorized Weight Quantizationcited by 15
  36. 2018ChannelNets: Compact and Efficient Convolutional Neural Networks via Channel-Wise Convolutionscited by 10
  37. 2018Building Efficient Deep Neural Networks with Unitary Group Convolutionscited by 9
  38. 2019NAIS: Neural Architecture and Implementation Search and its Applications in Autonomous Drivingcited by 9

50 most cited of 65 All 65 (JSON)

Built on by

Cited by year

Citing papers per year: 2021: 2; 2022: 3; 2023: 2; 2024: 3; 2025: 3.

13 papers.

Sort:

By area:

  1. 2022Hardware Approximate Techniques for Deep Neural Network Accelerators: A Surveycited by 39
  2. 2023Green Edge AI: A Contemporary Surveycited by 25
  3. 2024Learning from Students: Applying t-Distributions to Explore Accurate and Efficient Formats for LLMscited by 18
  4. 2021PhiNets: a scalable backbone for low-power AI at the edgecited by 16
  5. 2025Edge-Cloud Collaborative Computing on Distributed Intelligence and Model Optimization: A Surveycited by 14
  6. 2022Autonomy and Intelligence in the Computing Continuum: Challenges, Enablers, and Future Directions for Orchestrationcited by 12
  7. 2023FLIQS: One-Shot Mixed-Precision Floating-Point and Integer Quantization Searchcited by 6
  8. 2025MetaML-Pro: Cross-Stage Design Flow Automation for Efficient Deep Learning Accelerationcited by 4
  9. 20213U-EdgeAI: Ultra-Low Memory Training, Ultra-Low BitwidthQuantization, and Ultra-Low Latency Accelerationcited by 3
  10. 2024Radial Networks: Dynamic Layer Routing for High-Performance Large Language Modelscited by 1
  11. 2025Zero-Shot Visual Deepfake Detection: Can AI Predict and Prevent Fake Content Before It's Created?cited by 1
  12. 2022Compilation and Optimizations for Efficient Machine Learning on Embedded Systemscited by 0
1 more
  1. 2024Residual-INR: Communication Efficient On-Device Learning Using Implicit Neural Representationcited by 0