Research map / Architectures and efficiency / Neural architecture search
Constrained deep neural network architecture search for IoT devices accounting for hardware calibration
Florian Scheidegger, Luca Benini, Costas Bekas, A. Cristiano I. Malossi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Search narrow architecture spaces for floating-point models under the same memory constraints | Established IoT model · CIFAR-10 | accuracy: 70.64% -> 74.87% | Improved |
| Run models using 16-bit half types (ablation) | Narrow-space search · CIFAR-10 | accuracy: 74.87% -> 82.07% | Improved |
| Use model-optimized IEEE 754 reduced types (ablation) | Narrow-space search · CIFAR-10 | accuracy: 82.07% -> 83.45% | Improved |
| Run with half precision (ablation) | Trained models in the IoT-relevant domain under a given memory constraint · CIFAR-10 | accuracy: over 7% points | Improved |
| Choose an individual reduced-precision format for each model (ablation) | Trained models in the IoT-relevant domain under a given memory constraint · CIFAR-10 | accuracy: over 1% points further | Improved |
Closest papers by meaning
- SpArSe: Sparse Architecture Search for CNNs on Resource-Constrained Microcontrollers (NeurIPS 2019 · similarity 0.91)
- MCUNet: Tiny Deep Learning on IoT Devices (NeurIPS 2020 · similarity 0.89)
- Entropy-Driven Mixed-Precision Quantization for Deep Network Design (NeurIPS 2022 · similarity 0.88)
- TinyTTA: Efficient Test-time Adaptation via Early-exit Ensembles on Edge Devices (NeurIPS 2024 · similarity 0.86)
- TinyTrain: Resource-Aware Task-Adaptive Sparse Training of DNNs at the Data-Scarce Edge (ICML 2024 · similarity 0.86)
- AtomNet: Designing Tiny Models from Operators Under Extreme MCU Constraints (AAAI 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
