Skip to main content
BenchmarkComputer Visionv2017

COCO Detection

by Lin et al. / Microsoft · open-source · Last verified 2026-03-17

COCO Detection is the standard benchmark for object detection and instance segmentation, featuring 330,000 images with over 1.5 million annotated instances across 80 object categories. Mean Average Precision (mAP) at various IoU thresholds is the primary metric.

https://cocodataset.org
A
AGreat
Adoption: A+Quality: A+Freshness: BCitations: A+Engagement: F

Specifications

License
CC BY 4.0
Pricing
open-source
Capabilities
evaluation, object-detection, instance-segmentation
Integrations
Use Cases
model-evaluation, computer-vision, robotics
API Available
No
Evaluated Models
dino-detr, grounding-dino, co-detr, internimage-xl
Metrics
map, map-50, map-75
Methodology
Models evaluated on the 2017 val set (5,000 images). mAP averaged over IoU thresholds from 0.5 to 0.95 in steps of 0.05. mAP@50 and mAP@75 reported separately for comparison.
Last Run
2026-01-10
Tags
object-detection, instance-segmentation, vision, map, coco
Added
2026-03-17
Completeness
100%

Index Score

80.2
Adoption
95
Quality
90
Freshness
60
Citations
97
Engagement
0

Explore the full AI ecosystem on Agents as a Service