no code implementations • 7 Aug 2024 • Inas Bachiri, Hadjer Benmeziane, Smail Niar, Riyadh Baghdadi, Hamza Ouarnoughi, Abdelkrime Aries
Two notable techniques employed to achieve this goal are Hardware-aware Neural Architecture Search (HW-NAS) and Automatic Code Optimization (ACO).
Hardware Aware Neural Architecture Search Neural Architecture Search +1
no code implementations • 20 Feb 2024 • Halima Bouzidi, Smail Niar, Hamza Ouarnoughi, El-Ghazali Talbi
Our SONATA has seen up to sim$93. 6% Pareto dominance over the native NSGA-II, further stipulating the importance of self-adaptive evolution operators in HW-aware NAS.
Evolutionary Algorithms Hardware Aware Neural Architecture Search +1
1 code implementation • 12 Nov 2023 • Sofiane Bouaziz, Hadjer Benmeziane, Youcef Imine, Leila Hamdad, Smail Niar, Hamza Ouarnoughi
In fall detection using the MobiAct dataset, FLASH-RL outperforms FedAVG by up to 2. 82% in model's performance and reduces latency by up to 34. 75%.
no code implementations • 20 Sep 2023 • Hadjer Benmeziane, Kaoutar El Maghraoui, Hamza Ouarnoughi, Smail Niar
The mathematical instructions are then used as the basis for searching and selecting efficient replacement operators that maintain the accuracy of the original model while reducing computational complexity.
Hardware Aware Neural Architecture Search Neural Architecture Search
1 code implementation • 12 Sep 2023 • Mohamed Imed Eddine Ghebriout, Halima Bouzidi, Smail Niar, Hamza Ouarnoughi
In this paper, we propose Harmonic-NAS, a framework for the joint optimization of unimodal backbones and multimodal fusion networks with hardware awareness on resource-constrained devices.
General Classification Multimodal Text and Image Classification +1
no code implementations • 16 Jul 2023 • Mohanad Odema, Halima Bouzidi, Hamza Ouarnoughi, Smail Niar, Mohammad Abdullah Al Faruque
To achieve this, MaGNAS employs a two-tier evolutionary search to identify optimal GNNs and mapping pairings that yield the best performance trade-offs.
1 code implementation • 17 May 2023 • Hadjer Benmeziane, Corey Lammie, Irem Boybat, Malte Rasch, Manuel Le Gallo, Hsinyu Tsai, Ramachandran Muralidhar, Smail Niar, Ouarnoughi Hamza, Vijay Narayanan, Abu Sebastian, Kaoutar El Maghraoui
Digital processors based on typical von Neumann architectures are not conducive to edge AI given the large amounts of required data movement in and out of memory.
no code implementations • 10 May 2023 • Hadjer Benmeziane, Halima Bouzidi, Hamza Ouarnoughi, Ozcan Ozturk, Smail Niar
Deep learning has enabled various Internet of Things (IoT) applications.
no code implementations • 8 Mar 2023 • Lotfi Abdelkrim Mecharbat, Hadjer Benmeziane, Hamza Ouarnoughi, Smail Niar
Vision Transformers have enabled recent attention-based Deep Learning (DL) architectures to achieve remarkable results in Computer Vision (CV) tasks.
Ranked #1 on Image Classification on Visual Wake Words
Hardware Aware Neural Architecture Search Image Classification +3
no code implementations • 24 Feb 2023 • Halima Bouzidi, Mohanad Odema, Hamza Ouarnoughi, Smail Niar, Mohammad Abdullah Al Faruque
Heterogeneous MPSoCs comprise diverse processing units of varying compute capabilities.
1 code implementation • 6 Dec 2022 • Halima Bouzidi, Mohanad Odema, Hamza Ouarnoughi, Mohammad Abdullah Al Faruque, Smail Niar
Dynamic neural networks (DyNNs) have become viable techniques to enable intelligence on resource-constrained edge devices while maintaining computational efficiency.
no code implementations • 22 Jan 2021 • Hadjer Benmeziane, Kaoutar El Maghraoui, Hamza Ouarnoughi, Smail Niar, Martin Wistuba, Naigang Wang
Arguably their most significant impact has been in image classification and object detection tasks where the state of the art results have been obtained.
Hardware Aware Neural Architecture Search Image Classification +4
no code implementations • 21 Oct 2020 • Halima Bouzidi, Hamza Ouarnoughi, Smail Niar, Abdessamad Ait El Cadi
In this paper, we present and compare five (5) of the widely used Machine Learning based methods for execution time prediction of CNNs on two (2) edge GPU platforms.