AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Audio-Visual Segmentation with Semantics

Jinxing Zhou, Xuyang Shen, Jianyuan Wang, Jiayi Zhang, Weixuan Sun, Jing Zhang, Stan T. Birchfield, Dan Guo, Lingpeng Kong, Meng Wang, Yiran Zhong

IJCV 2024 · Vision-language and multimodal › Cross-modal learning · approach: Audio-visual source localization · 120 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 TPAVI audio-visual module (ablation)AVS (ResNet50) · AVSBench S4mIoU (%): 70.12 -> 72.79Improved
add TPAVI audio-visual module (ablation)AVS (PVT-v2) · AVSBench MS3mIoU (%): 48.21 -> 53.06Improved
add TPAVI audio-visual module (ablation)AVS (PVT-v2) · AVSBench AVSSmIoU (%): 27.71 -> 29.77Improved
add audio by direct addition of features (no TPAVI) (ablation)AVS (PVT-v2) · AVSBench S4mIoU (%): 77.76 -> 77.65No clear effect
add audio by direct addition of features (no TPAVI) (ablation)AVS (PVT-v2) · AVSBench MS3mIoU (%): 48.21 -> 51.55Improved
use TPAVI instead of direct audio-visual addition (ablation)AVS (ResNet50) · AVSBench S4mIoU (%): 70.54 -> 72.79Improved
add L_AVM-AV regularization to BCE (ablation)AVS (PVT-v2) · AVSBench MS3mIoU (%): 53.06 -> 54.00Improved
add L_AVM-AV regularization to BCE (ablation)AVS (ResNet50) · AVSBench AVSSmIoU (%): 18.88 -> 20.18Improved
add L_AVM-VV instead of L_AVM-AV (ablation)AVS (PVT-v2) · AVSBench MS3mIoU (%): 53.06 -> 53.77Improved
apply TPAVI at all four stages instead of stage 4 only (ablation)AVS (PVT-v2) · AVSBench MS3mIoU (%): 49.01 -> 54.00Improved
apply TPAVI at all four stages instead of best single stage (ablation)AVS (ResNet50) · AVSBench AVSSmIoU (%): 19.29 -> 20.18Improved
pretrain on Single-source subset (ablation)AVS (ResNet50) · AVSBench MS3mIoU (%): 47.88 -> 54.33Improved

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