mlsanityData to 14 Oct 2026

Smart at what cost? Characterising Mobile Deep Neural Networks in the wild

arXiv 2109.13963 ↗First posted 28 Sep 2021cs.LG cs.PFcited by 25

Not on the map. Why?

Abstract (arXiv)

With smartphones' omnipresence in people's pockets, Machine Learning (ML) on mobile is gaining traction as devices become more powerful. With applications ranging from visual filters to voice assistants, intelligence on mobile comes in many forms and facets. However, Deep Neural Network (DNN) inference remains a compute intensive workload, with devices struggling to support intelligence at the cost of responsiveness.On the one hand, there is significant research on reducing model runtime requirements and supporting deployment on embedded devices. On the other hand, the strive to maximise the accuracy of a task is supported by deeper and wider neural networks, making mobile deployment of state-of-the-art DNNs a moving target. In this paper, we perform the first holistic study of DNN usage in the wild in an attempt to track deployed models and match how these run on widely deployed devices. To this end, we analyse over 16k of the most popular apps in the Google Play Store to characterise their DNN usage and performance across devices of different capabilities, both across tiers and generations. Simultaneously, we measure the models' energy footprint, as a core cost dimension of any mobile deployment. To streamline the process, we have developed gaugeNN, a tool that automates the deployment, measurement and analysis of DNNs on devices, with support for different frameworks and platforms. Results from our experience study paint the landscape of deep learning deployments on smartphones and indicate their popularity across app developers. Furthermore, our study shows the gap between bespoke techniques and real-world deployments and the need for optimised deployment of deep learning models in a highly dynamic and heterogeneous ecosystem.

Key facts. Cited by 25 papers in the map (data through 14 Oct 2026). Builds on 36 of its 68 references that are in the map. First posted on arXiv 28 Sep 2021.

Builds on

36 of 68 references.

References · 36

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  3. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
  4. 2016Densely Connected Convolutional Networkscited by 6,720
  5. 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,859
  6. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,845
  7. 2014Sequence to Sequence Learning with Neural Networkscited by 2,986
  8. 2015Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Codingcited by 2,329
  9. 2017Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inferencecited by 1,593
  10. 2018MnasNet: Platform-Aware Neural Architecture Search for Mobilecited by 943
  11. 2019Towards Federated Learning at Scale: System Designcited by 832
  12. 2018SNIP: Single-shot Network Pruning based on Connection Sensitivitycited by 812
24 more
  1. 2017ICNet for Real-Time Semantic Segmentation on High-Resolution Imagescited by 261
  2. 2015Quantized Convolutional Neural Networks for Mobile Devicescited by 215
  3. 2021FjORD: Fair and Accurate Federated Learning under heterogeneous targets with Ordered Dropoutcited by 192
  4. 2020SPINN: Synergistic Progressive Inference of Neural Networks over Device and Cloudcited by 112
  5. 2021Federated Evaluation and Tuning for On-Device Personalization: System Design & Applicationscited by 85
  6. 2019AI Benchmark: All About Deep Learning on Smartphones in 2019cited by 77
  7. 2021Adaptive Inference through Early-Exit Networks: Design, Challenges and Directionscited by 75
  8. 2019BlazeFace: Sub-millisecond Neural Face Detection on Mobile GPUscited by 74
  9. 2019FasterSeg: Searching for Faster Real-time Semantic Segmentationcited by 73
  10. 2018A First Look at Deep Learning Apps on Smartphonescited by 47
  11. 2017FSSD: Feature Fusion Single Shot Multibox Detectorcited by 43
  12. 2020Mind Your Weight(s): A Large-scale Study on Insufficient Machine Learning Model Protection in Mobile Appscited by 36
  13. 2019An Empirical Study towards Characterizing Deep Learning Development and Deployment across Different Frameworks and Platformscited by 35
  14. 2019On-Device Neural Net Inference with Mobile GPUscited by 32
  15. 2020HAPI: Hardware-Aware Progressive Inferencecited by 30
  16. 2019EmBench: Quantifying Performance Variations of Deep Neural Networks across Modern Commodity Devicescited by 21
  17. 2021It's always personal: Using Early Exits for Efficient On-Device CNN Personalisationcited by 18
  18. 2021DynO: Dynamic Onloading of Deep Neural Networks from Cloud to Devicecited by 18
  19. 2019Neural Network Inference on Mobile SoCscited by 17
  20. 2020Scaling Up Online Speech Recognition Using ConvNetscited by 16
  21. 2018Latency and Throughput Characterization of Convolutional Neural Networks for Mobile Computer Visioncited by 13
  22. 2020Searching for Winograd-aware Quantized Networkscited by 11
  23. 2018A Family of Droids -- Android Malware Detection via Behavioral Modeling: Static vs Dynamic Analysiscited by 2
  24. 2018A Study of WhatsApp Usage Patterns and Prediction Models without Message Contentcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 8; 2023: 6; 2024: 7; 2025: 2; 2026: 2.

25 papers.

Sort:

  1. 2022EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformerscited by 116
  2. 2024MELTing point: Mobile Evaluation of Language Transformerscited by 36
  3. 2022NIPQ: Noise proxy-based Integrated Pseudo-Quantizationcited by 35
  4. 2023Mobile Foundation Model as Firmwarecited by 25
  5. 2024PocketLLM: Enabling On-Device Fine-Tuning for Personalized LLMscited by 18
  6. 2022Benchmarking of DL Libraries and Models on Mobile Devicescited by 12
  7. 2022The Future of Consumer Edge-AI Computingcited by 12
  8. 2023Green Federated Learningcited by 8
  9. 2023Maestro: Uncovering Low-Rank Structures via Trainable Decompositioncited by 8
  10. 2022Boosting DNN Cold Inference on Edge Devicescited by 6
  11. 2023Sparse-DySta: Sparsity-Aware Dynamic and Static Scheduling for Sparse Multi-DNN Workloadscited by 6
  12. 2022FedorAS: Federated Architecture Search under system heterogeneitycited by 4
13 more
  1. 2023Exploring the Performance and Efficiency of Transformer Models for NLP on Mobile Devicescited by 4
  2. 2024An exploration of the effect of quantisation on energy consumption and inference time of StarCoder2cited by 4
  3. 2022Edge Security: Challenges and Issuescited by 3
  4. 2023MultiTASC: A Multi-Tenancy-Aware Scheduler for Cascaded DNN Inference at the Consumer Edgecited by 3
  5. 2024A Picture is Worth 500 Labels: A Case Study of Demographic Disparities in Local Machine Learning Models for Instagram and TikTokcited by 2
  6. 2022NAWQ-SR: A Hybrid-Precision NPU Engine for Efficient On-Device Super-Resolutioncited by 1
  7. 2024CARIn: Constraint-Aware and Responsive Inference on Heterogeneous Devices for Single- and Multi-DNN Workloadscited by 1
  8. 2025CoFormer: Collaborating with Heterogeneous Edge Devices for Scalable Transformer Inferencecited by 1
  9. 2025Optimizing Storage Overhead of User Behavior Log for ML-embedded Mobile Appscited by 1
  10. 2026EdgeFlowerTune: Evaluating Federated LLM Fine-Tuning Under Realistic Edge System Constraintscited by 1
  11. 2024Dynamic DNNs and Runtime Management for Efficient Inference on Mobile/Embedded Devicescited by 0
  12. 2024MultiTASC++: A Continuously Adaptive Scheduler for Edge-Based Multi-Device Cascade Inferencecited by 0
  13. 2026Optimizing Feature Extraction for On-device Model Inference with User Behavior Sequencescited by 0