AI Paper Atlas

Research map / Detection and segmentation / CNN features

YOLACT: Real-Time Instance Segmentation

Daniel Bolya, Chong Zhou, Fanyi Xiao, Yong Jae Lee

ICCV 2019 · Detection and segmentation › CNN features · approach: Boundary-aware instance segmentation · 2,646 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (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 changedWhereMeasuredEffect
add Fast NMS in place of standard NMS (ablation)YOLACT-550 · COCO val2017mask AP: 30.0 -> 29.9No clear effect
add Fast NMS in place of standard NMS (ablation)YOLACT-550 · COCO val2017FPS: 24.0 -> 33.5Improved
add Fast NMS in place of standard NMS (ablation)Mask R-CNN · COCO val2017mask AP: 36.1 -> 35.8Hurt
add semantic segmentation loss during training (ablation)YOLACT-550 · COCOmask AP: +0.4 mAPImproved
increase number of prototypes from 8 to 32 (ablation)YOLACT-550 · COCO val2017mask AP: 26.8 -> 27.7Improved
increase prototypes from 32 to 64 (ablation)YOLACT-550 · COCO val2017mask AP: 27.7 -> 27.8No clear effect
increase prototypes from 64 to 256 (ablation)YOLACT-550 · COCO val2017mask AP: 27.8 -> 27.7No clear effect
add prototype plus coefficient mask assembly instead of fc mask output (ablation)one-stage detector · COCO val2017mask AP: 20.7 -> 29.9Improved
increase image size from 400 to 550 (ablation)YOLACT · COCO test-devmask AP: 24.9 -> 29.8Improved
increase image size from 550 to 700 (ablation)YOLACT · COCO test-devmask AP: 29.8 -> 31.2Improved
use ResNet-101 instead of ResNet-50 backbone (ablation)YOLACT-550 · COCO test-devmask AP: 28.2 -> 29.8Improved
use DarkNet-53 instead of ResNet-50 backbone (ablation)YOLACT-550 · COCO test-devmask AP: 28.2 -> 28.7Improved

Later papers that compared with this method

5 later accepted papers report YOLACT-700 as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy