CogSENet: Blind Image Deblurring with Blur-Conditioned Semantic Routing and Explicit Frequency Fusion
2026-06-29 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors address the challenge of fixing blurry images, especially when the blurring varies across the picture and is caused by unknown factors. They created CogSENet, a system inspired by how eagles see, using techniques that focus on important parts of the image based on meaning (semantics) and breaking the image into different detail levels to improve clarity. Their method also estimates how the blur changes across the image and uses this information along with semantic cues to adaptively restore sharpness. Experiments show that their approach works better than current methods while being simpler and also helps with related tasks like removing haze or noise.
blind image deblurringspatially varying blursemantic awarenesstoken regroupingwavelet transformfrequency decompositionblur field estimationCLIP semantic priorslatent feature modulation
Authors
Pan Wang, Yihao Hu, Xiujin Liu
Abstract
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.