Research map / Architectures and efficiency / CNN design
High Performance Depthwise and Pointwise Convolutions on Mobile Devices
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Apply ARM-specific loop rescheduling and register tiling to DWConv | DWConv · DWConv extracted from MobileNetV1, MobileNetV2, and MnasNet-A1 | speedup over TVM: 1.4× and up to 5.5× | Helped |
| Use RTRD loop blocking to keep output blocks in registers across the input-channel loop | PWConv · PWConv extracted from MobileNetV1, MobileNetV2, and MnasNet-A1 | speedup over TVM: 1.05× to 2.11× | Helped |
| Partition DWConv work across channels rather than output rows | DWConv · MobileNetV1, 4 threads | speedup: 2.2× to 3.9× | Helped |
| Use the proposed PWConv implementation with 4 threads | PWConv · MobileNetV1, 4 threads | speedup: 3.2× to 3.9× | Helped |
Closest papers by meaning
- MobileODE: An Extra Lightweight Network (NeurIPS 2025 · similarity 0.89)
- ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices (CVPR 2018 · similarity 0.86)
- RepViT: Revisiting Mobile CNN From ViT Perspective (CVPR 2024 · similarity 0.85)
- IFORMER: INTEGRATING CONVNET AND TRANSFORMER FOR MOBILE APPLICATION (ICLR 2025 · similarity 0.83)
- MobileMamba: Lightweight Multi-Receptive Visual Mamba Network (CVPR 2025 · similarity 0.83)
- FreeNet: Liberating Depth-Wise Separable Operations for Building Faster Mobile Vision Architectures (AAAI 2025 · similarity 0.82)
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.