AI Paper Atlas

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

Gaseous Object Detection

Kailai Zhou, Yibo Wang, Tao Lv, Qiu Shen, Xun Cao

TPAMI 2024 · Detection and segmentation › Open-set and few-shot detection · approach: Open-set and unknown object detection

Read the paper

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
replace frame-level Faster RCNN with video-level concatenation (ablation)Faster RCNN · GOD-VideoAP50: 36.15 -> 37.34Improved
add VSF_data to the concatenation baseline (ablation)video-level Faster RCNN · GOD-VideoAP50: 37.34 -> 43.88Improved
add VSF_fea on top of VSF_data (ablation)VSF RCNN · GOD-VideoAP50: 43.88 -> 51.08Improved
apply VSF in deeper stages (stages 1-4 vs none) (ablation)VSF RCNN · GOD-VideoAP50: 43.88 -> 51.08Improved
add temporal-axis shift to spatial W,H shift (ablation)VSF · GOD-VideoAP50: 48.18 -> 51.08Improved
use initial bias set +-1 +-2 instead of +-1 +-2 +-3 (ablation)VSF · GOD-VideoAP: 20.31 -> 20.43Improved
increase input frames from 8 to 16 (ablation)VSF RCNN · GOD-VideoAP: 20.43 -> 20.40No clear effect
increase input frames from 1 to 8 (ablation)VSF RCNN · GOD-VideoAP50: 36.15 -> 51.08Improved
add VSF to RetinaNetRetinaNet · GOD-VideoAP50: 37.13 -> 50.35Improved
add VSF to DETRDETR · GOD-VideoAP50: 26.41 -> 40.96Improved
add optical flow input to MOC (ablation)MOC · GOD-VideoAP50: 36.81 -> 34.50Hurt
use DETR-based STMixer instead of DETRDETR · GOD-VideoAP50: 26.41 -> 32.60Improved

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