AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Zero-Shot Quantization for Object Detectors using Off-the-Shelf Generative Models

Hyunho Lee, Kyomin Hwang, Hyeonjin Kim, Suyoung Kim, Sunghyun Wee, Nojun Kwak

ECCV 2026 · Detection and segmentation › Open-set and few-shot detection · approach: Zero-shot segmentation and detection · 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 Information-Dense Prompting to diffusion generation (ablation)GoodQ (YOLOv5-s) · MS-COCO val W4A4mAP / mAP50: 16.0 / 28.8 -> 19.3 / 33.8Improved
add Information-Dense Prompting to diffusion generation (ablation)GoodQ (YOLOv5-s) · MS-COCO val W8A8mAP / mAP50: 33.5 / 52.1 -> 34.3 / 53.5Improved
add Intrinsic Distribution-Aware Selection to diffusion generation (ablation)GoodQ (YOLOv5-s) · MS-COCO val W4A4mAP / mAP50: 16.0 / 28.8 -> 19.8 / 34.4Improved
combine IDP and IDAS (ablation)GoodQ (YOLOv5-s) · MS-COCO val W4A4mAP / mAP50: 19.3 / 33.8 -> 21.0 / 36.1Improved
replace one-hot pseudo-label detect loss with TANR (ablation)GoodQ (YOLOv5-s) · MS-COCO val W8A8mAP / mAP50: 33.9 / 52.8 -> 35.0 / 53.6Improved
replace one-hot pseudo-label detect loss with TANR (ablation)GoodQ (YOLOv5-s) · MS-COCO val W4A4mAP / mAP50: 20.6 / 35.8 -> 21.0 / 36.1Improved
increase image pool from 2k to 160k (ablation)GoodQ (YOLOv5-s) · MS-COCO val W3A3mAP / mAP50: 7.76 / 15.2 -> 10.20 / 19.7Improved
increase image pool from 2k to 160k (ablation)GoodQ (YOLOv5-s) · MS-COCO val W8A8mAP / mAP50: 34.3 / 53.5 -> 35.0 / 53.6No clear effect
select images by IDAS instead of random (ablation)GoodQ (YOLOv5-s) · MS-COCO val W4A4mAP / mAP50: 19.3 / 33.8 -> 21.0 / 36.1Improved
select images with most bounding boxes instead of random (ablation)GoodQ (YOLOv5-s) · MS-COCO val W4A4mAP / mAP50: 19.3 / 33.8 -> 17.5 / 30.7Hurt
use generated diffusion data instead of noise-optimized data (TSOD) at low bitYOLOv11-s QAT · MS-COCO val W3A3mAP: 0.5 -> 14.2Improved
use GoodQ instead of TSOD on Mask R-CNN Swin-TMask R-CNN Swin-T · MS-COCO val W3A3mAP / mAP50: 16.5 / 29.5 -> 21.6 / 36.7Improved

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