For ResNet-101 Faster R-CNN on COCO, sequential box refinement, global context, and multi-scale testing improve both mAP@0.5 and mAP@[.5,.95]. On test-dev, the single multi-scale model reaches 55.7% and 34.9%, and a three-network ensemble reaches 59.0% and 37.4%, winning the COCO 2015 detection task.

For ResNet-101 Faster R-CNN on COCO, sequential box refinement, global context, and multi-scale testing improve both mAP@0.5 and mAP@[.5,.95]. On test-dev, the single multi-scale model reaches 55.7% and 34.9%, and a three-network ensemble reaches 59.0% and 37.4%, winning the COCO 2015 detection task. · CiteArk

Measurement results

0/19 assessed

0 supported0 disputed or conflicting19 not assessed

The complete result group stays with its claim. Assessed rows come first; their judgments do not replace the whole-claim conclusion. Expand a row for conditions, sources, and evidence history.

Each row is a measurement with its reported value, latest available observation, and cumulative evidence state.
Reported	Latest observation	Evidence judgment
41.5%	—	Not assessed
21.2%	—	Not assessed
48.4%	—	Not assessed
27.2%	—	Not assessed
49.9%	—	Not assessed
29.9%	—	Not assessed
51.1%	—	Not assessed
30%	—	Not assessed
53.8%	—	Not assessed
32.5%	—	Not assessed
53.3%	—	Not assessed
32.2%	—	Not assessed

Reported and observed measurements

Assessments (0)

Runs (0)

Measurement results