AI Paper Atlas

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

Towards Open Vocabulary Learning: A Survey

Jianzong Wu, Xiangtai Li, Shilin Xu, Haobo Yuan, Henghui Ding, Yibo Yang, Xia Li, Jiangning Zhang, Yunhai Tong, Xudong Jiang, Bernard Ghanem, Dacheng Tao

TPAMI 2024 · Detection and segmentation › Open-set and few-shot detection · approach: Open-vocabulary object detection · 221 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
scale backbone from RN50-FPN to RN50x64F-VLM · LVIS 1.0 open-vocab detectionmask AP_r: 18.6 -> 32.8Improved
scale backbone from RN50 to RN50x64DST-Det · V3Det open-vocab detectionAP novel: 7.2 -> 13.5Improved
scale backbone from ViT-B/16 to ViT-L/16CLIPSelf · LVIS 1.0 open-vocab detectionmask AP_r: 25.3 -> 34.9Improved
replace RN50-FPN with Swin-B backboneDetic · LVIS 1.0 open-vocab detectionmask AP_r: 19.5 -> 23.9Improved
replace RN50-FPN with Swin-B backboneVLDet · LVIS 1.0 open-vocab detectionmask AP_r: 21.7 -> 26.3Improved
scale ViT-B/16 to ViT-L/14Open-VCLIP · UCF101 open-vocab video classificationaccuracy: 83.4 -> 87.6Improved
replace R50 with SwinB backboneMindVLT · LV-VIS valmAP_all: 14.1 -> 21.4Improved
use MindVLT instead of Detic-OWTB trackingDetic-OWTB (R50) · LV-VIS valmAP novel: 5.3 -> 9.4Improved
add Objaverse-LVIS to OpenShape training dataOpenShape-PointBERT · Objaverse-LVIS zero-shot 3D classificationTop1 accuracy: 39.1 -> 46.8Improved
train on OpenShape ensembled data instead of ShapeNet onlyOpenShape-PointBERT · Objaverse-LVIS zero-shot 3D classificationTop1 accuracy: 10.8 -> 39.1Improved
add language supervision (PLA training)PLA w/o training · ScanNet open-world 3D seg (B15/N4)hIoU: 39.7 -> 65.3Improved
replace PLA with RegionPLC (regional language supervision)PLA · ScanNet open-world 3D seg (B15/N4)hIoU: 65.3 -> 69.9Improved

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