# Instance Segmentation on MS COCO 2018 (test-dev)

Source: [Instance Segmentation on MS COCO 2018 (test-dev)](<https://www.sota2.com/research/sota/instance-segmentation-on-ms-coco-2018-test-dev>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:22:38.473Z

Evaluation of Instance Segmentation performance on the test-dev split of the MS COCO 2018 dataset, reporting standard COCO AP metrics across various IoU thresholds and object sizes.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Instance Segmentation |
| Dataset (including any stated split) | MS COCO 2018 (test-dev) |
| Metric | AP |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [AP](<https://www.sota2.com/research/sota/instance-segmentation-on-ms-coco-2018-test-dev.md?metric=AP>): Higher is better; unit: PERCENT.

- [AP @ IoU=0.50](<https://www.sota2.com/research/sota/instance-segmentation-on-ms-coco-2018-test-dev.md?metric=AP%20%40%20IoU%3D0.50>): Higher is better; unit: PERCENT.

- [AP @ IoU=0.75](<https://www.sota2.com/research/sota/instance-segmentation-on-ms-coco-2018-test-dev.md?metric=AP%20%40%20IoU%3D0.75>): Higher is better; unit: PERCENT.

- [AP (Small objects)](<https://www.sota2.com/research/sota/instance-segmentation-on-ms-coco-2018-test-dev.md?metric=AP%20(Small%20objects)>): Higher is better; unit: PERCENT.

- [AP (Medium objects)](<https://www.sota2.com/research/sota/instance-segmentation-on-ms-coco-2018-test-dev.md?metric=AP%20(Medium%20objects)>): Higher is better; unit: PERCENT.

- [AP (Large objects)](<https://www.sota2.com/research/sota/instance-segmentation-on-ms-coco-2018-test-dev.md?metric=AP%20(Large%20objects)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | AP | Result date | Report paper |
| --- | --- | --- | --- |
| Mask R-CNN w/ CARAFE | 35.9 | Not recorded | 68fa66911e209131cea10f20 |
| Mask R-CNN | 34.6 | Not recorded | 68fa66911e209131cea10f20 |

### Mask R-CNN w/ CARAFE

AP: 35.9. Result ID: 698cf8288b517eb4331e0df0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66911e209131cea10f20.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66911e209131cea10f20.md>)

[Implementation](<https://github.com/open-mmlab/mmdetection>)

Recorded evaluation configuration: {"Backbone":"ResNet-50"}. This record does not assert independent reproduction.

### Mask R-CNN

AP: 34.6. Result ID: 698cf8288b517eb4331e0def.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66911e209131cea10f20.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63f21e209131cea01b22.md>)

[Implementation](<https://github.com/facebookresearch/Detectron>)

Recorded evaluation configuration: {"Backbone":"ResNet-50"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.