2024

  • W. Hu, P. Henderson, J. Cano ‘DQA: An Efficient Method for Deep Quantization of Deep Neural Network Activations.’, to appear in Machine Learning with new Compute Paradigms Workshop (MLNCP) co-located with NeurIPS, Vancouver, Canada. [Paper]

  • P. Gibson, J. Cano, E. J. Crowley, A. Storkey, M. O’Boyle ‘DLAS: A Conceptual Model for Across-Stack Deep Learning Acceleration’, to appear in ACM Transaction on Architecture and Code Optimization (TACO). [Paper]

  • J. Haris, R. Saha, W. Hu, J. Cano ‘Designing Efficient LLM Accelerators for Edge Devices’, in Workshop on New Approaches for Addressing the Computing Requirements of LLMs and GNSs (ARC-LG), co-located with ISCA, Buenos Aires, Argentina. [Paper]

  • J. Haris, N. Bohm Agostini, A. Tumeo, D. Kaeli, J. Cano ‘Data Transfer Optimizations for Host-CPU and Accelerators in AXI4MLIR’, in 5th Compilers for Machine Learning Workshop (C4ML), co-located with CGO, Edinburgh, UK. [Paper]

  • R. Saha, J. Haris, J. Cano ‘Accelerating PoT Quantization on Edge Devices’, to appear in 31st IEEE International Conference on Electronics Circuits and Systems (ICECS). [Paper]

  • V. Sharma, D. Pau and J. Cano ‘Efficient Tiny Machine Learning for Human Activity Recognition on Low-Power Edge Devices’, to appear in IEEE Research and Technologies for Society and Industry (RTSI). [Paper]

  • N. Bohm Agostini, J. Haris, P. Gibson, M. Jawaweera, N. Rubin, Antonio Tumeo, J. L. Abellán, J. Cano, D. Kaeli ‘AXI4MLIR: User-Driven Automatic Host Code Generation for Custom AXI-Based Accelerators’, in IEEE/ACM International Symposium on Code Generation and Optimization (CGO). [Paper] [arXiv]

2023

  • P. Gibson ‘Compiler-centric Across-stack Deep Learning Acceleration’, Ph.D. Thesis, University of Glasgow. [Thesis]

  • N. Louloudakis, P. Gibson, J. Cano, and A. Rajan ‘DeltaNN: Assessing the Impact of Computational Environment Parameters on the Performance of Image Recognition Models’, to appear in IEEE International Conference on Software Maintenance and Evolution (ICSME). [Paper] [arXiv]

  • W. Hu, P. Gibson, and J. Cano, ‘ICE-Pick: Iterative Cost-Efficient Pruning for DNNs’, in Neural Compression Workshop (NCW), co-located with ICML, Honolulu, Hawaii, USA. [Paper]

  • N. Louloudakis, P. Gibson, J. Cano, and A. Rajan ‘Fault Localization for Buggy Deep Learning Framework Conversions in Image Recognition’, to appear in IEEE/ACM International Conference on Automated Software Engineering (ASE). [Paper] [arXiv]

  • F. Ayaz, I. Zakariyya, J. Cano, S. L. Keoh, J. Singer, D. Pau, M. Kharbouche-Harrari, ‘Improving Robustness Against Adversarial Attacks with Deeply Quantized Neural Networks’, in International Joint Conference on Neural Networks (IJCNN), Gold Coast, Australia, June 2023. [Pre-print]

  • J. Haris, P. Gibson, J. Cano, N. Bohm Agostini, D. Kaeli, ‘SECDA-TFLite: A Toolkit for Efficient Development of FPGA-based DNN Accelerators for Edge Inference’, in Elsevier Journal of Parallel and Distributed Computing (JPDC), Volume 173, March 2023. [Paper] [Code]

2022

  • N. Louloudakis, P. Gibson, J. Cano, A. Rajan, ‘Assessing Robustness of Image Recognition Models to Changes in the Computational Environment’, in NeurIPS ML Safety Workshop (MLSW) co-located with NeurIPS, Hybrid Conference, November-December 2022. [Pre-print]

  • P. Gibson, J. Cano, ‘Transfer-Tuning: Reusing Auto-Schedules for Efficient Tensor Program Code Generation’, in 31st International Conference on Parallel Architectures and Compilation Techniques (PACT), Chicago, USA, October 2022. [Paper] [arXiv] [Code artifact]

  • A. Stjerngren, P. Gibson, J. Cano, ‘Bifrost: End-to-End Evaluation and Optimization of Reconfigurable DNN Accelerators’, in IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS), Singapore, May 2022. [Paper] [arXiv] [Code]

  • P. Gibson, J. Cano, ‘Productive Reproducible Workflows for DNNs: A Case Study for Industrial Defect Detection’, in 4th Workshop on Accelerated Machine Learning (AccML) co-located with HiPEAC, Budapest, Hungary, June 2022. [Paper]

2021

  • S. Dong, Y. Sun, N. Bohm Agostini, E. Karimi, D. Lowell, J. Zhou, J. Cano, J. L. Abellán, D. Kaeli, ‘Spartan: A Sparsity-Adaptive Framework to Accelerate Deep Neural Network Training on GPUs’, in IEEE Transactions on Parallel and Distributed Systems (TPDS), Volume 32, Issue 10, October 2021. [Paper]

  • J. Haris, P. Gibson, J. Cano, N. B. Agostini, and D. Kaeli, ‘SECDA: Efficient Hardware/Software Co-Design of FPGA-based DNN Accelerators for Edge Inference’, in 2021 IEEE 33rd International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD), Belo Horizonte, Brazil, October 2021. [Paper] [arXiv] [Code]

  • M. Lofqvist, J. Cano, ‘Optimizing Data Processing in Space for Object Detection in Satellite Imagery’, in 35th Annual Small Satellite Conference (SmallSat), Virtual Event, August 2021. [Paper] [arXiv]

2020

  • N. Bohm Agostini, S. Dong, E. Karimi, M. Torrents, J. Cano, J. L. Abellán, D. Kaeli, ‘Design Space Exploration of Accelerators and End-to-End DNN Evaluation with TFLITE-SOC’, in 32nd IEEE International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD), Porto, Portugal, September 2020. [Paper]

  • P. Gibson and J. Cano, ‘Orpheus: A new deep learning framework for easy deployment and evaluation of edge inference’, in 2020 IEEE international symposium on performance analysis of systems and software (ISPASS), Virtual Meeting, August 2020. [Paper] [arXiv]

  • M. Lofqvist, J. Cano, ‘Accelerating Deep Learning Applications in Space’, in 34th Annual Small Satellite Conference (SmallSat), Virtual Event, August 2020. [Paper] [arXiv]

  • P. Gibson, J. Cano, J. Turner, E. J. Crowley, M. O’Boyle, and A. Storkey, ‘Optimizing grouped convolutions on edge devices’, in 2020 IEEE 31st international conference on application-specific systems, architectures and processors (ASAP), Manchester, UK, July 2020. [Paper] [arXiv] [Code]