MnasFPN: Learning Latency-aware Pyramid Architecture for Object Detection on Mobile Devices
arXiv:1912.01106
Abstract
Despite the blooming success of architecture search for vision tasks in resource-constrained environments, the design of on-device object detection architectures have mostly been manual. The few automated search efforts are either centered around non-mobile-friendly search spaces or not guided by on-device latency. We propose MnasFPN, a mobile-friendly search space for the detection head, and combine it with latency-aware architecture search to produce efficient object detection models. The learned MnasFPN head, when paired with MobileNetV2 body, outperforms MobileNetV3+SSDLite by 1.8 mAP at similar latency on Pixel. It is also both 1.0 mAP more accurate and 10% faster than NAS-FPNLite. Ablation studies show that the majority of the performance gain comes from innovations in the search space. Further explorations reveal an interesting coupling between the search space design and the search algorithm, and that the complexity of MnasFPN search space may be at a local optimum.
10 pages, 7 figures
References in corpus (13)
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
- MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
- EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
- Neural Architecture Search with Reinforcement Learning
- R-FCN: Object Detection via Region-based Fully Convolutional Networks
- SGDR: Stochastic Gradient Descent with Warm Restarts
- AMC: AutoML for Model Compression and Acceleration on Mobile Devices
- Efficient Neural Architecture Search via Parameter Sharing
- Random Search and Reproducibility for Neural Architecture Search
- DetNAS: Backbone Search for Object Detection
- Exploring Randomly Wired Neural Networks for Image Recognition
- ThunderNet: Towards Real-time Generic Object Detection
- NAS-FCOS: Fast Neural Architecture Search for Object Detection