AI Paper Atlas

Research map / Detection and segmentation / CNN features

Inlier-Centric Post-Training Quantization for Object Detection Models

Minsu Kim, Dongyeun Lee, Jaemyung Yu, Jiwan Hur, Giseop Kim, Junmo Kim

ICLR 2026 · Detection and segmentation › CNN features · approach: Region-based and query-based detectors · 0 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 heatmap top-K selection to anomaly-only quantization (ablation)InlierQ · COCO 2D camera W4A4mAP: 32.5 -> 34.5Improved
add heatmap top-K selection to anomaly-only quantization (ablation)InlierQ · nuScenes 3D camera W4A4mAP: 24.8 -> 25.8Improved
add heatmap top-K selection to anomaly-only quantization (ablation)InlierQ · nuScenes 3D LiDAR W4A4mAP: 44.2 -> 45.7Improved
restrict quantization to inliers instead of anomalies (with heatmap) (ablation)InlierQ · COCO 2D camera W4A4mAP: 34.5 -> 34.7Improved
exclude anomalies from inlier set (ablation)InlierQ · COCO 2D camera W4A4mAP: 34.6 -> 34.7Improved
exclude anomalies from inlier set (ablation)InlierQ · nuScenes 3D camera W4A4mAP: 26.1 -> 26.4Improved
exclude anomalies from inlier set (ablation)InlierQ · nuScenes 3D LiDAR W4A4mAP: 46.0 -> 46.6Improved
use EM instead of K-Means for inlier-anomaly clustering (ablation)InlierQ · nuScenes 3D LiDAR W4A4mAP: 43.5 -> 46.6Improved
use EM instead of SVM for inlier-anomaly clustering (ablation)InlierQ · nuScenes 3D LiDAR W4A4mAP: 46.6 -> 46.6No clear effect
raise heatmap top-K beyond training-time K (ablation)InlierQ · nuScenes 3D detectionmAPHurt
use InlierQ instead of LiDAR-PTQ at W4A4LiDAR-PTQ · nuScenes 3D LiDAR CenterPointmAP: 39.5 -> 46.6Improved
use InlierQ instead of BRECQ at 8-bit weights and activationsBRECQ · COCO RetinaNet W8A8mAP: 38.9 -> 39.0No clear effect

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