Research

For the most updated overview of my papers, please visit my scholar page.

Paper pre-prints

  • A. Kouris*, S. I. Venieris*, S. Laskaridis, N. D. Lane, “Fluid Batching: Exit-Aware Preemptive Serving of Early-Exit Neural Networks on Edge NPUs”, 2022 [arXiv:2209.13443]
  • L. Dudziak*, S. Laskaridis*, J. Marques-Fernandez*, “FedorAS: Federated Architecture Search under system heterogeneity”, 2022 [arXiv:2206.11239]

Peer-reviewed papers

  • A. Kouris, S. I. Venieris*, S. Laskaridis*, N. D. Lane, “Multi-Exit Semantic Segmentation Networks”, European Conference in Computer Vision (ECCV) and International Conference on Mobile Systems, Applications and Services (MobiSys), 2022 [preprint, short paper]
  • M. Almeida*, S. Laskaridis*, S. I. Venieris*, I. Leontiadis*, N. D. Lane, “DynO: Dynamic Onloading of Deep Neural Networks from Cloud to Device”, Special Issue on Accelerating AI on the Edge in Transactions of ACM Embedded Computing Systems (TECS), 2022 [paper, preprint]
  • S. Horvath*, S. Laskaridis*, M. Almeida*, et al., “FjORD: Fair and Accurate Federated Learning under heterogeneous targets with Ordered Dropout”, Conference on Neural Information Processing Systems (NeurIPS), 2021, Spotlight (top 3%) [paper, poster]
  • M. Almeida*, S. Laskaridis*, A. Mehrotra, et al., Smart at what cost? Characterising Mobile Deep Neural Networks in the wild, ACM Internet Measurement Conference (IMC), 2021 [paper, preprint, video]
  • S. Laskaridis, D Spathis, M Almeida, Federated mobile sensing for activity recognition, International Conference on Mobile Computing and Networking (MobiCom), 2021 [paper]
  • S. Laskaridis, A. Kouris, N. D. Lane, “Adaptive Inference through Early-Exit Networks: Design, Challenges and Directions”, International Workshop on Embedded and Mobile Deep Learning (EMDL) at MobiSys, 2021 [paper, preprint]
  • I. Leontiadis*, S. Laskaridis*, S. I. Venieris*, N. D. Lane, “It’s always personal: Using Early Exits for Efficient On-Device CNN Personalisation”, International Workshop on Mobile Computing Systems and Applications (HotMobile), 2021 [paper, preprint]
  • S. Laskaridis*, S. I. Venieris*, M. Almeida*, et al., “SPINN: Synergistic Progressive Inference of Neural Networks over Device and Cloud”, International Conference on Mobile Computing and Networking (MobiCom), 2020 [paper, preprint, video]
  • S. Laskaridis*, S. I. Venieris*, H. Kim, N. D. Lane, “HAPI: Hardware-Aware Progressive Inference”, International Conference on Computer-Aided Design (ICCAD), 2020 [paper, preprint]
  • M. Almeida*, S. Laskaridis*, I. Leontiadis*, S. Venieris*, N. D. Lane, “EmBench: Demystifying The Performance Of Deep Neural Networks Across Modern Commodity Devices”, International Workshop on Embedded and Mobile Deep Learning (EMDL) at MobiSys, 2019 [paper, preprint]
  • Ł. Dudziak, M. S. Abdelfattah, R. Vipperla, S. Laskaridis, N. D. Lane, “ShrinkML: End-to-End ASR Model Compression Using Reinforcement Learning”, INTERSPEECH, 2019 [paper, preprint]
  • S. Laskaridis, V. Bahyl et al., “Tape SCSI monitoring and encryption at CERN”, Journal of Physics Conference Series (JPCS), CHEP, 2016 - Work also presented in Open Compute Project - Archival storage [paper, talk]

Patents

  • S. Laskaridis, S. Horvath, M. Almeida, I.Leontiadis, S. I. Venieris, “Method, system and apparatus for federated learning”, US Patent App., No.17586178
  • M.Almeida, S. Laskaridis, S. I. Venieris, I. Leontiadis, “Method and system for neural network execution distribution”, US Patent App., No.17420259
  • S. Laskaridis, H. Kim, S. I. Venieris, “Method and system for implementing a variable accuracy neural network”, US Patent App., No.16923447

Talks

  • Panelist at PhD Symposium, ICDCS 2022 - Panelists: Mirco Musolesi (UCL), Giovanni Pau (University of Bologna)
  • “Measuring and tackling system heterogeneity in on-device DNN deployment”, University of Avignon and FAIR Menlo Park
  • “Federated Learning: Current landscape and challenges”, Federated Mobile Sensing for Activity Recognition Tutorial, MobiCom 2021 [video]
  • Panel co-ordinator on “Federated Learning and Distributed Learning”, DistributedML 2021 [video] - Panelists: Vijay Janapa Reddi (Harvard University, MLCommons), Bo Li (UIUC), Mosharaf Chowdhury (University of Michigan), Tianyi Chen (RPI)
  • “Fair and Accurate Federated Learning under heterogeneous targets with Ordered Dropout”, FLOW Seminar [video]
  • “Is the future of IoT learning Decentralised”, MobiSys 2021 IoT Day [link]
  • “Federated Learning with Ordered Dropout for system heterogeneity”, Flower Summit 2021 [video]
  • Panel co-ordinator on “Federated Learning and Collaborative Inference”, DistributedML 2020 [video] - Panelists: Dimitris Papailiopoulos (Wisconsin-Madison), Amanda Prorok (University of Cambridge), Fahim Kawsar (Bell Labs, TU Delft), Ilias Leontiadis (Samsung AI)
  • “SPINN: Synergistic Progressive Inference of Neural Networks over Device and Cloud”, Mobicom 2021 [video]

© 2022. All rights reserved.

Powered by Hydejack v9.1.6