Rethinking Channel Dimensions for Efficient Model Design
| Autor: | Dongyoon Han, Sangdoo Yun, Youngjoon Yoo, Byeongho Heo |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
FOS: Computer and information sciences
Network architecture Computer science business.industry Computer Vision and Pattern Recognition (cs.CV) Rank (computer programming) Computer Science - Computer Vision and Pattern Recognition Parameterized complexity Object detection Piecewise linear function Dimension (vector space) Computer engineering Artificial intelligence business Transfer of learning Communication channel |
| Zdroj: | CVPR |
| DOI: | 10.1109/cvpr46437.2021.00079 |
| Popis: | Designing an efficient model within the limited computational cost is challenging. We argue the accuracy of a lightweight model has been further limited by the design convention: a stage-wise configuration of the channel dimensions, which looks like a piecewise linear function of the network stage. In this paper, we study an effective channel dimension configuration towards better performance than the convention. To this end, we empirically study how to design a single layer properly by analyzing the rank of the output feature. We then investigate the channel configuration of a model by searching network architectures concerning the channel configuration under the computational cost restriction. Based on the investigation, we propose a simple yet effective channel configuration that can be parameterized by the layer index. As a result, our proposed model following the channel parameterization achieves remarkable performance on ImageNet classification and transfer learning tasks including COCO object detection, COCO instance segmentation, and fine-grained classifications. Code and ImageNet pretrained models are available at https://github.com/clovaai/rexnet. Comment: 13 pages, 8 figures, CVPR 2021 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|