Alapfogalmak
Proposing a novel framework, EMIP, for Video Camouflaged Object Detection that explicitly handles motion cues through an interactive prompting mechanism.
Kivonat
EMIP introduces a two-stream architecture to address camouflaged segmentation and optical flow estimation simultaneously. The model incorporates a frozen pre-trained optical flow fundamental model and utilizes an interactive prompting scheme inspired by emerging visual prompt learning. By integrating segmentation-to-motion and motion-to-segmentation prompts, EMIP achieves state-of-the-art results on popular VCOD benchmarks.
Statisztikák
EMIP achieves new state-of-the-art records on popular VCOD benchmarks.
Notable margins of improvement (∼17.0%/5.5% average) over the previous best model SLT-Net.
Idézetek
"EMIP effectively leverages noise-robust motion to detect and segment video camouflaged objects."
"Prompt learning strategy for the motion stream can better exploit its potential on limited VCOD data."