CogSENet: Blind Image Deblurring with Blur-Conditioned Semantic Routing and Explicit Frequency Fusion

Hugging Face Daily Papers Papers
blind-image-deblurring semantic-aware frequency-fusion eagle-vision state-space-model wavelet-transform clip-priors

Summary

CogSENet introduces a blind image deblurring framework inspired by eagle vision, using semantic-aware modules and frequency decomposition to improve restoration quality and structural fidelity, outperforming state-of-the-art methods.

Blind image deblurring demands the recovery of high-fidelity details and coherent structures from complex, unknown degradations. Current blind image deblurring methods struggle with real-world, spatially varying degradations, and lack the semantic awareness necessary to reliably differentiate valid textures from artifacts. To bridge this gap, we propose CogSENet, a dynamic, semantic-aligned reconstruction framework inspired by the eagle's visual system. By mimicking the eagle's active saccadic scanning, we devise a Semantic-Driven State Space Module (SDSSM) with semantic-aware token regrouping via differentiable routing, enabling prompt-conditioned long-range dependency modeling. To ensure physically interpretable recovery of textures and structures, a BiFreqFusionBlock (BFFB) mirrors functional differentiation of the eagle's retina by decomposing features into high and low frequencies using wavelet transforms. Finally, we estimate a continuous Blur Field (CBF) from blur image and fuse it with CLIP semantic priors to modulate the deepest latent features, emulating focal adaptation and enabling adaptive restoration under spatially non-uniform blur. Extensive experiments demonstrate that CogSENetoutperforms state-of-the-art deblurring methods in both visual quality and structural fidelity with fewer parameters, while also performing favorably on dehazing, deraining, and denoising tasks.
Original Article
View Cached Full Text

Cached at: 07/02/26, 03:49 PM

Paper page - CogSENet: Blind Image Deblurring with Blur-Conditioned Semantic Routing and Explicit Frequency Fusion

Source: https://huggingface.co/papers/2606.30030

Abstract

CogSENet presents a novel blind image deblurring framework inspired by eagle vision, incorporating semantic-aware modules and frequency decomposition for improved restoration quality and structural fidelity.

Blind image deblurringdemands the recovery of high-fidelity details and coherent structures from complex, unknown degradations. Currentblind image deblurringmethods struggle with real-world, spatially varying degradations, and lack the semantic awareness necessary to reliably differentiate valid textures from artifacts. To bridge this gap, we propose CogSENet, a dynamic, semantic-aligned reconstruction framework inspired by the eagle’s visual system. By mimicking the eagle’s active saccadic scanning, we devise a Semantic-Driven State Space Module (SDSSM) withsemantic-aware token regroupingviadifferentiable routing, enabling prompt-conditionedlong-range dependency modeling. To ensure physically interpretable recovery of textures and structures, aBiFreqFusionBlock(BFFB) mirrors functional differentiation of the eagle’s retina by decomposing features into high and low frequencies usingwavelet transforms. Finally, we estimate acontinuous Blur Field(CBF) from blur image and fuse it withCLIP semantic priorsto modulate the deepest latent features, emulatingfocal adaptationand enabling adaptive restoration underspatially non-uniform blur. Extensive experiments demonstrate that CogSENetoutperforms state-of-the-art deblurring methods in both visual quality and structural fidelity with fewer parameters, while also performing favorably on dehazing, deraining, and denoising tasks.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2606\.30030

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.30030 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.30030 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.30030 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Generated Contents Enrichment

Hugging Face Daily Papers

This paper proposes a joint adversarial framework that uses graph convolutional networks to enrich sparse scene graphs before generating semantically richer visual content. The approach makes scene enrichment explicit at the representation level, producing visually plausible and structurally coherent images.