Research map / Detection and segmentation / CNN features
MobileDets: Searching for Object Detection Architectures for Mobile Accelerators
Check 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 |
|---|---|---|---|
| search the detector architecture directly for COCO and CPU latency | MobileDet search on CPU · COCO, CPU | mAP: +1.7 mAP | Helped |
| add regular-convolution blocks to the search space (ablation) | MobileDet on EdgeTPU · COCO, EdgeTPU | mAP: 0.8 mAP improvement | Helped |
| add regular-convolution blocks to the search space (ablation) | MobileDet on DSP · COCO, DSP | mAP: 26.9 mAP @ 12.2 ms to 28.5 mAP @ 11.9ms | Helped |
| add regular convolutions to the searched architecture (ablation) | MobileDet on CPU · COCO, CPU | mAP | No clear effect |
| search the detector architecture directly for COCO and EdgeTPU latency | MobileDet · COCO, EdgeTPU | mAP: +3.7 mAP | Helped |
| search the detector architecture directly for COCO and DSP latency | MobileDet · COCO, DSP | mAP: +3.2 mAP | Helped |
| add regular-convolution blocks to the search space (ablation) | MobileDet on CPU · COCO, CPU | mAP | No clear effect |
| add regular-convolution blocks to the search space (ablation) | MobileDet on EdgeTPU · COCO, EdgeTPU | mAP | Helped |
| add regular-convolution blocks to the search space (ablation) | MobileDet on DSP · COCO, DSP | mAP | Helped |
| add Tucker layers to IBN+Fused search space (ablation) | MobileDet on CPU · COCO, CPU | mAP: 22.7 -> 23.7 | Helped |
| add Tucker layers to IBN+Fused search space (ablation) | MobileDet on EdgeTPU · COCO, EdgeTPU | mAP: 25.4 -> 25.5 | Mixed |
| add Tucker layers to IBN+Fused search space (ablation) | MobileDet on DSP · COCO, DSP | mAP: 28.5 -> 28.5 | No clear effect |
Closest papers by meaning
- MnasFPN: Learning Latency-Aware Pyramid Architecture for Object Detection on Mobile Devices (CVPR 2020 · similarity 0.89)
- MnasNet: Platform-Aware Neural Architecture Search for Mobile (CVPR 2019 · similarity 0.88)
- Mobile Video Object Detection With Temporally-Aware Feature Maps (CVPR 2018 · similarity 0.88)
- MobileNetV2: Inverted Residuals and Linear Bottlenecks (CVPR 2018 · similarity 0.88)
- Can't Slow Me Down: Learning Robust and Hardware-Adaptive Object Detectors against Latency Attacks for Edge Devices (CVPR 2025 · similarity 0.85)
- IFORMER: INTEGRATING CONVNET AND TRANSFORMER FOR MOBILE APPLICATION (ICLR 2025 · similarity 0.85)
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.