Awesome Split Computing & Early Exit Papers

🏠 About

Here is a curated list of Split Computing & Early Exit papers and resources.

🚧 This is an active repository; you can watch for the latest advances.

The ✨ indicates that this is an article that I co-authored.

If you find it useful, please kindly star ⭐ this repository.

📢 Latest Updates

[2025-05-18]: Made this repository public.

❤️ Community Support

This repository is maintained by Luigi Capogrosso (luigi.capogrosso@univr.it). I welcome feedback, suggestions, and contributions that can help improve this repository to make it a valuable resource that benefits the entire community.

I will actively maintain this repository by incorporating new research as it emerges. If you have any suggestions regarding the taxonomy, find any missed papers, or update any pre-print arXiv paper that has been accepted to some venue, please email me or submit a pull request using the following markdown format.

Paper Title, <ins>Conference/Journal/Preprint</ins>, [[article](link)], [[other resources](link)].

📌 Table of Contents

Articles are sorted by year, from most recent to oldest.

2025 | 2024 | 2023 | 2022 | 2021 | 2020 | 2019 | 2018 | 2017 | 2016

2025

• ✨ LO-SC: Local-Only Split Computing for Accurate Deep Learning on Edge Devices, International Conference on VLSI Design (VLSID), [article], [website].

• A Multi-task Supervised Compression Model for Split Computing, Winter Conference on Applications of Computer Vision (WACV), [article].

• ✨ Controllers for Edge-Cloud Cyber-Physical Systems, International Conference on COMmunication Systems and NETworks (COMSNETS), [article], [website].

• LVMScissor: Split and Schedule Large Vision Model Inference on Mobile Edges via Salp Swarm Algorithm, IEEE Transactions on Mobile Computing, [article].

• Request Deadline Split and Interference-Aware Request Migration in Edge Cloud, Concurrency and Computation: Practice and Experience, [article].

2024

• ✨ MTL-Split: Multi-Task Learning for Edge Devices using Split Computing, Design Automation Conference (DAC), [article], [website].

• Neuromorphic Split Computing With Wake-Up Radios: Architecture and Design via Digital Twinning, IEEE Transactions on Signal Processing, [article].

• ✨ Enhancing Split Computing and Early Exit Applications through Predefined Sparsity, Forum on Specification & Design Languages (FDL), [article], [website].

• Enhancing the Reliability of Split Computing Deep Neural Networks, International Symposium on On-Line Testing and Robust System Design (IOLTS), [article].

• ✨ Learning-Enabled CPS for Edge-Cloud Computing, International Symposium on Industrial Embedded Systems (SIES), [article], [website].

• Smart Split: Leveraging TinyML and Split Computing for Efficient Edge AI, Symposium on Edge Computing (SEC), [article].

• Split DNN Inference for Exploiting Near-Edge Accelerators, International Conference on Edge Computing and Communications (EDGE), [article].

• A novel middleware for adaptive and efficient split computing for real-time object detection, Pervasive and Mobile Computing, [article].

• Fast and fair split computing for accelerating deep neural network (DNN) inference, ICT Express, [article].

2023

• ✨ Split-Et-Impera: A Framework for the Design of Distributed Deep Learning Applications, International Symposium on Design and Diagnostics of Electronic Circuits and Systems (DDECS), [article], [website].

• Dynamic Split Computing for Efficient Deep EDGE Intelligence, International Conference on Acoustics, Speech and Signal Processing (ICASSP), [article].

• PrivyNAS: Privacy-Aware Neural Architecture Search for Split Computing in Edge–Cloud Systems, IEEE Internet of Things Journal, [article].

• SplitEE: Early Exit in Deep Neural Networks with Split Computing, Third International Conference on AI-ML Systems (AIMLSystems), [article].

• SplitBeam: Effective and Efficient Beamforming in Wi-Fi Networks Through Split Computing, International Conference on Distributed Computing Systems (ICDCS), [article].

• Distributed Split Computing System in Cooperative Internet of Things (IoT), IEEE Access, [article].

• DNN Split Computing: Quantization and Run-Length Coding are Enough, Global Communications Conference, [article].

• Efficient Communication-Computation Tradeoff for Split Computing: A Multi-Tier Deep Reinforcement Learning Approach, Global Communications Conference, [article].

• SC2 Benchmark: Supervised Compression for Split Computing, Transactions on Machine Learning Research, [article].

2022

• ✨ I-SPLIT: Deep Network Interpretability for Split Computing, International Conference on Pattern Recognition (ICPR), [article], [website].

• BottleFit: Learning Compressed Representations in Deep Neural Networks for Effective and Efficient Split Computing, International Symposium on a World of Wireless, Mobile and Multimedia Networks (WoWMoM), [article].

• Supervised Compression for Resource-Constrained Edge Computing Systems, Winter Conference on Applications of Computer Vision (WACV), [article].

2021

• Packet-Loss-Tolerant Split Inference for Delay-Sensitive Deep Learning in Lossy Wireless Networks, Conference on Computer Vision and Pattern Recognition (CVPR), [article].

• A Splittable DNN-Based Object Detector for Edge-Cloud Collaborative Real-Time Video Inference, International Conference on Advanced Video and Signal Based Surveillance (AVSS), [article].

• Neural Compression and Filtering for Edge-assisted Real-time Object Detection in Challenged Networks, International Conference on Pattern Recognition (ICPR), [article].

• Cut, Distil and Encode (CDE): Split Cloud-Edge Deep Inference, International Conference on Sensing, Communication, and Networking (SECON), [article].

• Optimal Branch Location for Cost-effective Inference on Branchynet, International Conference on Big Data (Big Data), [article].

• Packet-Loss-Tolerant Split Inference for Delay-Sensitive Deep Learning in Lossy Wireless Networks, Global Communications Conference (GLOBECOM), [article].

• A Probabilistic Re-Intepretation of Confidence Scores in Multi-Exit Models, Entropy, [article].

• Zero Time Waste: Recycling Predictions in Early Exit Neural Networks, A dvances in Neural Information Processing Systems (NeurIPS), [article].

2020

• Head Network Distillation: Splitting Distilled Deep Neural Networks for Resource-Constrained Edge Computing Systems, IEEE Access, [article].

• Split Computing for Complex Object Detectors: Challenges and Preliminary Results, International Workshop on Embedded and Mobile Deep Learning, article].

• BottleNet++: An End-to-End Approach for Feature Compression in Device-Edge Co-Inference Systems, International Conference on Communications Workshops (ICC Workshops), [article].

• Resolution Adaptive Networks for Efficient Inference, Conference on Computer Vision and Pattern Recognition (CVPR), [article].

• Fast and Accurate Streaming CNN Inference via Communication Compression on the Edge, International Conference on Internet-of-Things Design and Implementation (IoTDI), [article].

• Joint Device-Edge Inference over Wireless Links with Pruning, International Workshop on Signal Processing Advances in Wireless Communications (SPAWC), [article].

• SPINN: synergistic progressive inference of neural networks over device and cloud, Annual International Conference on Mobile Computing and Networking (MobiCom), [article].

• Deep compressive offloading: speeding up neural network inference by trading edge computation for network latency, Conference on Embedded Networked Sensor Systems (SenSys), [article].

• Lightweight Compression Of Neural Network Feature Tensors For Collaborative Intelligence, International Conference on Multimedia and Expo (ICME), [article].

• CRIME: Input-Dependent Collaborative Inference for Recurrent Neural Networks, IEEE Transactions on Computers, [article].

• Back-And-Forth Prediction for Deep Tensor Compression, International Conference on Acoustics, Speech and Signal Processing (ICASSP), [article].

• Depth-Adaptive Transformer, International Conference on Learning Representations (ICLR), [article].

• Dual Dynamic Inference: Enabling More Efficient, Adaptive, and Controllable Deep Inference, IEEE Journal of Selected Topics in Signal Processing, [article].

• Early Exit or Not: Resource-Efficient Blind Quality Enhancement for Compressed Images, European Conference on Computer Vision(ECCV), [article].

2019

• JointDNN: An Efficient Training and Inference Engine for Intelligent Mobile Cloud Computing Services, IEEE Transactions on Mobile Computing, [article].

• BottleNet: A Deep Learning Architecture for Intelligent Mobile Cloud Computing Services, International Symposium on Low Power Electronics and Design (ISLPED), [article].

• Distilled Split Deep Neural Networks for Edge-Assisted Real-Time Systems, Workshop on Hot Topics in Video Analytics and Intelligent Edges, [article].

• DynExit: A Dynamic Early-Exit Strategy for Deep Residual Networks, International Workshop on Signal Processing Systems (SiPS), [article].

• Boomerang: On-Demand Cooperative Deep Neural Network Inference for Edge Intelligence on the Industrial Internet of Things, IEEE Network, [article].

• Exploiting Energy-Accuracy Trade-off through Contextual Awareness in Multi-Stage Convolutional Neural Networks, International Symposium on Quality Electronic Design (ISQED), [article].

• Distillation-Based Training for Multi-Exit Architectures, International Conference on Computer Vision (ICCV), [article].

2018

• Auto-tuning Neural Network Quantization Framework for Collaborative Inference Between the Cloud and Edge, Artificial Neural Networks and Machine Learning (ICANN), [article].

• Learning IoT in Edge: Deep Learning for the Internet of Things with Edge Computing, IEEE Network, [article].

• Computation Offloading for Machine Learning Web Apps in the Edge Server Environment, International Conference on Distributed Computing Systems (ICDCS), [article].

• Deep Feature Compression for Collaborative Object Detection, International Conference on Image Processing (ICIP), [article].

2017

• Neurosurgeon: Collaborative Intelligence Between the Cloud and Mobile Edge, ACM SIGARCH Computer Architecture News, [article].

• A Dynamic Deep Neural Network Design for Efficient Workload Allocation in Edge Computing, International Conference on Computer Design (ICCD), [article].

• Distributed Deep Neural Networks Over the Cloud, the Edge and End Devices, International Conference on Distributed Computing Systems (ICDCS), [article].

2016

• BranchyNet: Fast Inference via Early Exiting from Deep Neural Networks, International Conference on Pattern Recognition (ICPR), [article].