| Architecture | RPN, ASPP, Convolution, RFP, ReLU, SAC, Dense Connections, FPN, ResNet, RoIAlign |
|---|---|
| lr sched | 1x |
| SHOW MORE |
| Architecture | RPN, ASPP, Convolution, RFP, ReLU, Dense Connections, FPN, ResNet, RoIAlign |
|---|---|
| lr sched | 1x |
| SHOW MORE |
| Architecture | RPN, ASPP, Convolution, SAC, ReLU, Dense Connections, FPN, ResNet, RoIAlign |
|---|---|
| lr sched | 1x |
| SHOW MORE |
[ALGORITHM]
We provide the config files for DetectoRS: Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolution.
@article{qiao2020detectors,
title={DetectoRS: Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolution},
author={Qiao, Siyuan and Chen, Liang-Chieh and Yuille, Alan},
journal={arXiv preprint arXiv:2006.02334},
year={2020}
}
DetectoRS requires COCO and COCO-stuff dataset for training. You need to download and extract it in the COCO dataset path. The directory should be like this.
mmdetection
├── mmdet
├── tools
├── configs
├── data
│ ├── coco
│ │ ├── annotations
│ │ ├── train2017
│ │ ├── val2017
│ │ ├── test2017
| | ├── stuffthingmaps
DetectoRS includes two major components:
They can be used independently. Combining them together results in DetectoRS. The results on COCO 2017 val are shown in the below table.
| Method | Detector | Lr schd | Mem (GB) | Inf time (fps) | box AP | mask AP | Config | Download |
|---|---|---|---|---|---|---|---|---|
| RFP | Cascade + ResNet-50 | 1x | 7.5 | - | 44.8 | config | model | log | |
| SAC | Cascade + ResNet-50 | 1x | 5.6 | - | 45.0 | config | model | log | |
| DetectoRS | Cascade + ResNet-50 | 1x | 9.9 | - | 47.4 | config | model | log | |
| RFP | HTC + ResNet-50 | 1x | 11.2 | - | 46.6 | 40.9 | config | model | log |
| SAC | HTC + ResNet-50 | 1x | 9.3 | - | 46.4 | 40.9 | config | model | log |
| DetectoRS | HTC + ResNet-50 | 1x | 13.6 | - | 49.1 | 42.6 | config | model | log |
Note: This is a re-implementation based on MMDetection-V2. The original implementation is based on MMDetection-V1.
Object Detection on COCO minival
| MODEL | BOX AP |
|---|---|
| DetectoRS (HTC + ResNet-50, 1x) | 49.1 |
| DetectoRS (Cascade + ResNet-50, 1x) | 47.4 |
| DetectoRS RFP-only (HTC + ResNet-50, 1x) | 46.6 |
| SAC (HTC + ResNet-50, 1x) | 46.4 |
| SAC (Cascade + ResNet-50, 1x) | 45.0 |
| DetectoRS RFP-only (Cascade + ResNet-50, 1x) | 44.8 |
Instance Segmentation on COCO minival
| MODEL | MASK AP |
|---|---|
| DetectoRS (HTC + ResNet-50, 1x) | 42.6 |
| SAC (HTC + ResNet-50, 1x) | 40.9 |
| DetectoRS RFP-only (HTC + ResNet-50, 1x) | 40.9 |
