# CPUBone: Efficient Vision Backbone Design for Devices with Low Parallelization Capabilities

arXiv: https://arxiv.org/abs/2603.26425 · HTML: https://mlsanity.com/paper/2603.26425/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 37 references that are in the map.
- First posted on arXiv 27 Mar 2026.

## Builds on (most cited first, 25 of 25)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [SGDR: Stochastic Gradient Descent with Warm Restarts](https://mlsanity.com/paper/1608.03983/) (2016)
- [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://mlsanity.com/paper/1801.04381/) (2018)
- [MMDetection: Open MMLab Detection Toolbox and Benchmark](https://mlsanity.com/paper/1906.07155/) (2019)
- [EfficientNetV2: Smaller Models and Faster Training](https://mlsanity.com/paper/2104.00298/) (2021)
- [PVT v2: Improved Baselines with Pyramid Vision Transformer](https://mlsanity.com/paper/2106.13797/) (2021)
- [Panoptic Feature Pyramid Networks](https://mlsanity.com/paper/1901.02446/) (2019)
- [MetaFormer Is Actually What You Need for Vision](https://mlsanity.com/paper/2111.11418/) (2021)
- [YOLOv10: Real-Time End-to-End Object Detection](https://mlsanity.com/paper/2405.14458/) (2024)
- [Mobile-Former: Bridging MobileNet and Transformer](https://mlsanity.com/paper/2108.05895/) (2021)
- [FastViT: A Fast Hybrid Vision Transformer using Structural Reparameterization](https://mlsanity.com/paper/2303.14189/) (2023)
- [BiFormer: Vision Transformer with Bi-Level Routing Attention](https://mlsanity.com/paper/2303.08810/) (2023)
- [Run, Don't Walk: Chasing Higher FLOPS for Faster Neural Networks](https://mlsanity.com/paper/2303.03667/) (2023)
- [RepViT: Revisiting Mobile CNN From ViT Perspective](https://mlsanity.com/paper/2307.09283/) (2023)
- [Rethinking Vision Transformers for MobileNet Size and Speed](https://mlsanity.com/paper/2212.08059/) (2022)
- [MobileOne: An Improved One millisecond Mobile Backbone](https://mlsanity.com/paper/2206.04040/) (2022)
- [MobileNetV4 -- Universal Models for the Mobile Ecosystem](https://mlsanity.com/paper/2404.10518/) (2024)
- [SHViT: Single-Head Vision Transformer with Memory Efficient Macro Design](https://mlsanity.com/paper/2401.16456/) (2024)
- [GhostNetV2: Enhance Cheap Operation with Long-Range Attention](https://mlsanity.com/paper/2211.12905/) (2022)
- [MobileViG: Graph-Based Sparse Attention for Mobile Vision Applications](https://mlsanity.com/paper/2307.00395/) (2023)
- [Separable Self and Mixed Attention Transformers for Efficient Object Tracking](https://mlsanity.com/paper/2309.03979/) (2023)
- [Lightweight Vision Transformer with Bidirectional Interaction](https://mlsanity.com/paper/2306.00396/) (2023)
- [Efficient Modulation for Vision Networks](https://mlsanity.com/paper/2403.19963/) (2024)
- [LowFormer: Hardware Efficient Design for Convolutional Transformer Backbones](https://mlsanity.com/paper/2409.03460/) (2024)

## Abstract (arXiv)

Recent research on vision backbone architectures has predominantly focused on optimizing efficiency for hardware platforms with high parallel processing capabilities. This category increasingly includes embedded systems such as mobile phones and embedded AI accelerator modules. In contrast, CPUs do not have the possibility to parallelize operations in the same manner, wherefore models benefit from a specific design philosophy that balances amount of operations (MACs) and hardware-efficient execution by having high MACs per second (MACpS). In pursuit of this, we investigate two modifications to standard convolutions, aimed at reducing computational cost: grouping convolutions and reducing kernel sizes. While both adaptations substantially decrease the total number of MACs required for inference, sustaining low latency necessitates preserving hardware-efficiency. Our experiments across diverse CPU devices confirm that these adaptations successfully retain high hardware-efficiency on CPUs. Based on these insights, we introduce CPUBone, a new family of vision backbone models optimized for CPU-based inference. CPUBone achieves state-of-the-art Speed-Accuracy Trade-offs (SATs) across a wide range of CPU devices and effectively transfers its efficiency to downstream tasks such as object detection and semantic segmentation. Models and code are available at https://github.com/altair199797/CPUBone.