AI Paper Atlas

Research map / Detection and segmentation / CNN features

PointINS: Point-Based Instance Segmentation

Lu Qi, Yi Wang, Yukang Chen, Ying-Cong Chen, Xiangyu Zhang, Jian Sun, Jiaya Jia

TPAMI 2021 · Detection and segmentation › CNN features · approach: Boundary-aware instance segmentation · 35 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 9 feature channels to the instance-agnostic template (shape 9x16x16 instead of 1x16x16) (ablation)PointINS · COCO valmask AP: 30.4 -> 32.5Improved
use 9x16x16 template reshape instead of 1x48x48 natural representation (ablation)PointINS · COCO valmask AP: 29.6 -> 32.5Improved
increase template channels beyond 9 (25 channels) (ablation)PointINS · COCO valmask AP: 32.5 -> 32.5No clear effect
add explicit box information I^z for instance-aware weights instead of detection features (ablation)PointINS · COCO valmask AP: 27.9 -> 32.5Improved
add the ReLU after the FC weight predictor (ablation)PointINS · COCO valmask AP: 31.7 -> 32.5Improved
stack a second FC+ReLU in the weight predictor (ablation)PointINS · COCO valmask AP: 32.5 -> 32.6No clear effect
fuse template and weights by dynamic convolution instead of addition (ablation)PointINS · COCO valmask AP: 31.1 -> 32.5Improved
fuse template and weights by dynamic convolution instead of concatenation (ablation)PointINS · COCO valmask AP: 30.9 -> 32.5Improved
use 3x3 instead of 1x1 dynamic convolution kernel (ablation)PointINS · COCO valmask AP: 32.5 -> 32.5No clear effect
increase sampled positive points from 32 to 128 (ablation)PointINS · COCO valmask AP: 29.6 -> 32.5Improved
sample 256 positive points instead of 128 (ablation)PointINS · COCO valmask AP: 32.5 -> 32.4No clear effect
lower anchor IoU threshold from 0.5 to 0.4 (ablation)PointINS · COCO valmask AP: 32.5 -> 30.2Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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