Hybrid Convolution and Vision Transformer NAS Search Space for TinyML Image Classification
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917176692178944 |
|---|---|
| author | Djajapermana, Mikhael Reiber, Moritz Mueller-Gritschneder, Daniel Schlichtmann, Ulf |
| author_facet | Djajapermana, Mikhael Reiber, Moritz Mueller-Gritschneder, Daniel Schlichtmann, Ulf |
| contents | Hybrids of Convolutional Neural Network (CNN) and Vision Transformer (ViT) have outperformed pure CNN or ViT architecture. However, since these architectures require large parameters and incur large computational costs, they are unsuitable for tinyML deployment. This paper introduces a new hybrid CNN-ViT search space for Neural Architecture Search (NAS) to find efficient hybrid architectures for image classification. The search space covers hybrid CNN and ViT blocks to learn local and global information, as well as the novel Pooling block of searchable pooling layers for efficient feature map reduction. Experimental results on the CIFAR10 dataset show that our proposed search space can produce hybrid CNN-ViT architectures with superior accuracy and inference speed to ResNet-based tinyML models under tight model size constraints. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_02992 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Hybrid Convolution and Vision Transformer NAS Search Space for TinyML Image Classification Djajapermana, Mikhael Reiber, Moritz Mueller-Gritschneder, Daniel Schlichtmann, Ulf Computer Vision and Pattern Recognition Machine Learning Hybrids of Convolutional Neural Network (CNN) and Vision Transformer (ViT) have outperformed pure CNN or ViT architecture. However, since these architectures require large parameters and incur large computational costs, they are unsuitable for tinyML deployment. This paper introduces a new hybrid CNN-ViT search space for Neural Architecture Search (NAS) to find efficient hybrid architectures for image classification. The search space covers hybrid CNN and ViT blocks to learn local and global information, as well as the novel Pooling block of searchable pooling layers for efficient feature map reduction. Experimental results on the CIFAR10 dataset show that our proposed search space can produce hybrid CNN-ViT architectures with superior accuracy and inference speed to ResNet-based tinyML models under tight model size constraints. |
| title | Hybrid Convolution and Vision Transformer NAS Search Space for TinyML Image Classification |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2511.02992 |