@article{ART003399421},
author={In-Chul Han and KIM, SOO KYUN},
title={GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation},
journal={Journal of Internet of Things and Convergence},
issn={2466-0078},
year={2026},
volume={12},
number={5},
pages={18}
TY - JOUR
AU - In-Chul Han
AU - KIM, SOO KYUN
TI - GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation
JO - Journal of Internet of Things and Convergence
PY - 2026
VL - 12
IS - 5
PB - The Korea Internet of Things Society
SP - 18
EP -
SN - 2466-0078
AB - Camouflaged Object Segmentation (COS) aims to segment objects that are visually similar to their surrounding backgrounds. Due to subtle foreground-background differences, ambiguous boundaries, and irregular structures, COS requires stable integration of local details and high-level semantic cues. In this paper, we propose GAGS, an Adaptive Gaussian Aggregation and Guided Sampling framework. GAGS organizes hierarchical features into detail-oriented and semantic-oriented representations, complements them through frequency fusion, and converts the fused representation into a high-resolution feature through guided sampling. The restoration decoder then recovers fine object structures using shallow spatial cues and coarse-guided calibration. Rather than relying on fixed hierarchical fusion or repeated refinement, GAGS adaptively aggregates multi-resolution features according to their feature-level contributions. Experiments on CAMO, COD10K, and NC4K show competitive performance, achieving 0.855 and 0.953 on COD10K.
KW - Camouflaged Object Segmentation;Frequency Fusion;Adaptive Gaussian Mechanism;Guided Sampling;Multi-Scale Feature Integration
DO -
UR -
ER -
In-Chul Han and KIM, SOO KYUN. (2026). GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation. Journal of Internet of Things and Convergence, 12(5), 18.
In-Chul Han and KIM, SOO KYUN. 2026, "GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation", Journal of Internet of Things and Convergence, vol.12, no.5 18.
In-Chul Han, KIM, SOO KYUN "GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation" Journal of Internet of Things and Convergence 12.5 18 (2026) : 18.
In-Chul Han, KIM, SOO KYUN. GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation. 2026; 12(5), 18.
In-Chul Han and KIM, SOO KYUN. "GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation" Journal of Internet of Things and Convergence 12, no.5 (2026) : 18.
In-Chul Han; KIM, SOO KYUN. GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation. Journal of Internet of Things and Convergence, 12(5), 18.
In-Chul Han; KIM, SOO KYUN. GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation. Journal of Internet of Things and Convergence. 2026; 12(5) 18.
In-Chul Han, KIM, SOO KYUN. GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation. 2026; 12(5), 18.
In-Chul Han and KIM, SOO KYUN. "GAGS: Adaptive Gaussian Aggregation and Guided Sampling for Camouflaged Object Segmentation" Journal of Internet of Things and Convergence 12, no.5 (2026) : 18.
1