mlsanityData to 7 Oct 2026

Single-Path NAS: Designing Hardware-Efficient ConvNets in less than 4 Hours

arXiv 1904.02877 ↗First posted 5 Apr 2019cs.LG cs.CV stat.MLcited by 87

Not on the map. Why?

Abstract (arXiv)

Can we automatically design a Convolutional Network (ConvNet) with the highest image classification accuracy under the runtime constraint of a mobile device? Neural architecture search (NAS) has revolutionized the design of hardware-efficient ConvNets by automating this process. However, the NAS problem remains challenging due to the combinatorially large design space, causing a significant searching time (at least 200 GPU-hours). To alleviate this complexity, we propose Single-Path NAS, a novel differentiable NAS method for designing hardware-efficient ConvNets in less than 4 hours. Our contributions are as follows: 1. Single-path search space: Compared to previous differentiable NAS methods, Single-Path NAS uses one single-path over-parameterized ConvNet to encode all architectural decisions with shared convolutional kernel parameters, hence drastically decreasing the number of trainable parameters and the search cost down to few epochs. 2. Hardware-efficient ImageNet classification: Single-Path NAS achieves 74.96% top-1 accuracy on ImageNet with 79ms latency on a Pixel 1 phone, which is state-of-the-art accuracy compared to NAS methods with similar constraints (<80ms). 3. NAS efficiency: Single-Path NAS search cost is only 8 epochs (30 TPU-hours), which is up to 5,000x faster compared to prior work. 4. Reproducibility: Unlike all recent mobile-efficient NAS methods which only release pretrained models, we open-source our entire codebase at: https://github.com/dstamoulis/single-path-nas.

Key facts. Cited by 87 papers in the map (data through 7 Oct 2026). First posted on arXiv 5 Apr 2019.

Builds on

No reference list yet.

Built on by

Cited by year

Citing papers per year: 2021: 36; 2022: 20; 2023: 12; 2024: 11; 2025: 8.

87 papers.

Sort:

By area:

  1. 2021Sustainable AI: Environmental Implications, Challenges and Opportunitiescited by 242
  2. 2021HW-NAS-Bench:Hardware-Aware Neural Architecture Search Benchmarkcited by 83
  3. 2021A Comprehensive Survey on Hardware-Aware Neural Architecture Searchcited by 60
  4. 2023ZiCo: Zero-shot NAS via Inverse Coefficient of Variation on Gradientscited by 57
  5. 2024Flextron: Many-in-One Flexible Large Language Modelcited by 42
  6. 2021HR-NAS: Searching Efficient High-Resolution Neural Architectures with Lightweight Transformerscited by 39
  7. 2021Inspect, Understand, Overcome: A Survey of Practical Methods for AI Safetycited by 35
  8. 2022Machine Learning for Microcontroller-Class Hardware: A Reviewcited by 28
  9. 2021CompOFA: Compound Once-For-All Networks for Faster Multi-Platform Deploymentcited by 26
  10. 2021Pruning Self-attentions into Convolutional Layers in Single Pathcited by 24
  11. 2023Zero-Shot Neural Architecture Search: Challenges, Solutions, and Opportunitiescited by 21
  12. 2021HardCoRe-NAS: Hard Constrained diffeRentiable Neural Architecture Searchcited by 20
38 more
  1. 2021Bit-Mixer: Mixed-precision networks with runtime bit-width selectioncited by 20
  2. 2022CNN Filter DB: An Empirical Investigation of Trained Convolutional Filterscited by 20
  3. 2022Efficient Architecture Search for Diverse Taskscited by 18
  4. 2022PVNAS: 3D Neural Architecture Search with Point-Voxel Convolutioncited by 16
  5. 2024The 3D-PC: a benchmark for visual perspective taking in humans and machinescited by 16
  6. 2021Trilevel Neural Architecture Search for Efficient Single Image Super-Resolutioncited by 15
  7. 2021Searching by Generating: Flexible and Efficient One-Shot NAS with Architecture Generatorcited by 15
  8. 2021Layer Folding: Neural Network Depth Reduction using Activation Linearizationcited by 14
  9. 2021NetAdaptV2: Efficient Neural Architecture Search with Fast Super-Network Training and Architecture Optimizationcited by 13
  10. 2021Auto-NBA: Efficient and Effective Search Over the Joint Space of Networks, Bitwidths, and Acceleratorscited by 12
  11. 2022DropNAS: Grouped Operation Dropout for Differentiable Architecture Searchcited by 12
  12. 2021FLASH: Fast Neural Architecture Search with Hardware Optimizationcited by 11
  13. 2021Learning Versatile Neural Architectures by Propagating Network Codescited by 9
  14. 2021EEEA-Net: An Early Exit Evolutionary Neural Architecture Searchcited by 9
  15. 2021Searching for Efficient Multi-Stage Vision Transformerscited by 9
  16. 2024FL-NAS: Towards Fairness of NAS for Resource Constrained Devices via Large Language Modelscited by 9
  17. 2021One Proxy Device Is Enough for Hardware-Aware Neural Architecture Searchcited by 8
  18. 2022EAutoDet: Efficient Architecture Search for Object Detectioncited by 7
  19. 2024LLM-dCache: Improving Tool-Augmented LLMs with GPT-Driven Localized Data Cachingcited by 7
  20. 2021Single-path Bit Sharing for Automatic Loss-aware Model Compressioncited by 6
  21. 2021FNAS: Uncertainty-Aware Fast Neural Architecture Searchcited by 6
  22. 2021NAS-FCOS: Efficient Search for Object Detection Architecturescited by 6
  23. 2021Contrastive Self-supervised Neural Architecture Searchcited by 5
  24. 2025Efficient Quantum Convolutional Neural Networks for Image Classification: Overcoming Hardware Constraintscited by 5
  25. 2025L-SWAG: Layer-Sample Wise Activation with Gradients information for Zero-Shot NAS on Vision Transformerscited by 5
  26. 2021EH-DNAS: End-to-End Hardware-aware Differentiable Neural Architecture Searchcited by 4
  27. 2022RoHNAS: A Neural Architecture Search Framework with Conjoint Optimization for Adversarial Robustness and Hardware Efficiency of Convolutional and Capsule Networkscited by 4
  28. 2022Pareto-aware Neural Architecture Generation for Diverse Computational Budgetscited by 4
  29. 2024Boosting Order-Preserving and Transferability for Neural Architecture Search: a Joint Architecture Refined Search and Fine-tuning Approachcited by 4
  30. 2021Pareto-Frontier-aware Neural Architecture Generation for Diverse Budgetscited by 3
  31. 2021ZeBRA: Precisely Destroying Neural Networks with Zero-Data Based Repeated Bit Flip Attackcited by 3
  32. 2022Multi-Prior Learning via Neural Architecture Search for Blind Face Restorationcited by 3
  33. 2022Tackling Neural Architecture Search With Quality Diversity Optimizationcited by 3
  34. 2023FLORA: Fine-grained Low-Rank Architecture Search for Vision Transformercited by 3
  35. 2021CONet: Channel Optimization for Convolutional Neural Networkscited by 2
  36. 2021BINAS: Bilinear Interpretable Neural Architecture Searchcited by 2
  37. 2021RT-RCG: Neural Network and Accelerator Search Towards Effective and Real-time ECG Reconstruction from Intracardiac Electrogramscited by 2
  38. 2021Searching for TrioNet: Combining Convolution with Local and Global Self-Attentioncited by 2

50 most cited of 87 All 87 (JSON)