A frequency-aware compressed video quality enhancement framework that improves visual quality by adaptively enhancing high-frequency details and texture structures and reduces compression artifacts and enhances perceptual detail quality compared to existing approaches is proposed.
Abstract
Compressed video typically suffers from degraded visual quality due to artifacts such as blurring, blocking effects, and the attenuation of high-frequency details. These distortions make it challenging to accurately restore fine textures using conventional enhancement approaches that rely mainly on spatial domain information, as such methods lack effective modeling of high-frequency components degraded during compression. To address the above limitation, this paper proposes a frequency-aware compressed video quality enhancement framework that improves visual quality by adaptively enhancing high-frequency details and texture structures. The proposed method utilizes multiple consecutive frames and introduces a spatio-temporal feature alignment module to generate aligned features. A multi-scale high-frequency feature extraction strategy is then applied to capture fine details at different scales. Furthermore, an adaptive enhancement mechanism with spatial gating and high-frequency component weighting is proposed to selectively refine high-frequency details and suppress unnecessary enhancement. The frequency-enhanced features are then integrated with aligned features through a gated fusion strategy. Finally, the proposed quality enhancement module predicts a residual map, which is added to the center frame of the input sequence to produce the enhanced output. Experimental results demonstrate that the proposed method effectively reduces compression artifacts and enhances perceptual detail quality compared to existing approaches.
Existing image super-resolution methods still suffer from limitations in edge-structure restoration, high-frequency texture preservation, and artifact suppression, which may lead to blurred contours and unnatural textures. To address these issues, this paper proposes an image super-resolution method based on hierarchical feature aggregation and Laplacian high-frequency compensation. First, a Hierarchical Feature Aggregation Attention Block (HFAB) is designed in the generator to progressively extract image features at different levels through multiple convolutional layers. A High-Frequency Variance Adaptive Channel Attention Block (HFVB) is further introduced to adaptively enhance key texture and edge information. Second, a Laplacian Adaptive Upsampling (LAU) module is developed to combine low-frequency content reconstruction with high-frequency detail compensation, thereby strengthening edge contours, preserving fine textures, and reducing artifacts. Finally, a Dissimilarity Structural Similarity Index Measure (DSSIM) loss is incorporated into the loss function to constrain local structural consistency and further improve the structural preservation and perceptual quality of reconstructed images. Experimental results on Set5, Set14, BSD100, and Urban100 show that, compared with SRGAN, the proposed method improves PSNR by 0.31 dB, 0.18 dB, 0.14 dB, and 0.14 dB, respectively, while reducing LPIPS by 0.0234, 0.0183, 0.0222, and 0.0222. These results indicate that the proposed method provides consistent improvements over SRGAN and achieves modest, metric-dependent gains over ESRGAN, suggesting an incremental enhancement in reconstruction accuracy and perceptual quality on both natural image benchmarks and complex urban scene datasets.
Existing video motion magnification methods commonly decompose video frames into multiple spatial frequency bands using pyramid representations, such as the phase-based magnification method and the fast Riesz-pyramid approach. These pyramid-based decompositions, in combination with Gabor or Hilbert transforms, extract the instantaneous phase of images across different orientations. Motion magnification is then achieved by amplifying the phase variations between consecutive frames, followed by image reconstruction. However, due to the inherent limitations of the Gabor and Hilbert transforms, these approaches can only effectively capture instantaneous phase information in low- and mid-frequency ranges, while high-frequency components are not synchronously magnified. This mismatch often results in reconstruction artifacts.
To overcome this limitation, this paper proposes a wavelet-based video motion magnification framework. The proposed method employs a spatial Butterworth filter to decompose video frames into low-, mid-, and high-frequency bands, and leverages the wavelet transform to extract high-frequency phase information for synchronous magnification. By enhancing the representation of high-frequency components, the method effectively suppresses artifacts and improves visual fidelity. Experimental results demonstrate that the proposed approach achieves artifact-free motion magnification with superior accuracy and visual quality compared to existing methods.
Kun Xie, Liangliang Cheng· e-Journal of Nondestructive...· 0 citations
All-in-one image restoration has recently attracted considerable attention for its ability to address multiple degradation types within a single, unified framework. However, existing methods often incur substantial computational overhead, especially when incorporating explicit degradation priors via complex auxiliary branches, hindering their practical deployment. In this paper, we propose AdaptIR, an efficient all-in-one image restoration network equipped with adaptive frequency enhancement. Recognizing that different degradations impact distinct frequency subbands and exhibit spatially varying restoration demands, we design an Adaptive Frequency Enhancement Module (AFEM) that couples frequency learning with adaptive convolutions to better capture frequency-aware information. Specifically, AFEM learns pixel-wise adaptive attention weights to modulate the spectra of dynamic convolutions, enabling spatially adaptive and content-aware restoration. Furthermore, we introduce a lightweight backbone featuring a Receptive Field Expansion Module (RFEM), which enlarges the receptive field of a convolutional U-shaped architecture by convolving wavelet-transform coefficients. By integrating the plug-and-play AFEM into the bottleneck of the baseline model, AdaptIR achieves state-of-the-art performance on all-in-one image restoration tasks involving multiple degradations, while maintaining high computational efficiency. Moreover, the proposed model can be readily extended to single-degradation tasks (e.g., dehazing, desnowing, and deraining) and domain-specific applications, including ultra-high-definition (UHD), medical, and remote sensing image restoration.
This work proposes MDFI (Multi-Domain Features Integration), a compressed video quality enhancement approach that features a novel Frame-Prediction Feature Transform (FPFT) module to process prediction information to enhance decoded video quality.
Sang NguyenQuang, Hieu Bui Minh, BuiDinh Dang et al.· IEEE transactions on consume...· 0 citations
This paper addresses the challenge in low-light image and video enhancement often suffering from over-brightening, color distortion, structure degradation and inter-frame flickering, and presents an enhanced adaptive histogram specification (AHS) method to tackle the problem systematically with a balancing act of adaptive enhancement techniques. Driven by luminance distribution statistics, the AHS method organizes the enhancement process into a unified framework consisting of intensity-adaptive estimation, color preservation, structural rollback, and temporal smoothing. Specifically, (1) enhancement intensity is jointly estimated via the cumulative distribution function (CDF) distance and global luminance deviation, and local luminance correction is introduced to handle non-uniform illumination; (2) in the color space, the original and enhanced chrominance are adaptively fused according to saturation, and difference shrinkage is combined to suppress perceptible color casts; (3) a structural weight is used to perform conservative rollback for edge and texture regions during luminance fusion, reducing over-enhancement artifacts; (4) for video enhancement, temporal smoothing is applied to the enhancement intensity in the parameter domain and combined with a scene response mechanism to suppress flickers caused by frame-wise statistical jitter. Based on a preliminary dataset collected with the same device and a unified evaluation protocol, AHS achieves more balanced luminance preservation, color consistency, and structural fidelity across multiple static scenes, with mean AMBE=8.44, 𝛥𝐸94=3.76 and SSIM=0.884, significantly outperforming traditional methods like global histogram equalization, contrast limited adaptive histogram equalization (CLAHE) and adaptive gamma correction with weighting distribution (AGCWD). For video enhancement, experimental results show that AHS effectively reduces inter-frame fluctuations while maintaining low color difference, achieving a measure of Std |𝛥𝑌|=0.0969, a reduction by 58.86% and 81.62% as compared to CLAHE and AGCWD, respectively. The experimental results demonstrate that AHS can provide a more stable, interpretable, and controllable enhancement of images or videos under cross-scene conditions, offering a reproducible and promising technique for engineering deployment of low-light image and video enhancement.
Yifu Yang, Jianhua Xuan· International Conference on...· 0 citations
High-Compression videos suffer from severe distortions, among which degradation in person regions has the greatest impact on viewers’ immersive experience. Existing quality enhancement techniques usually focus on overall image denoising or super-resolution, often overlooking the crucial recovery of fine structures in these essential person regions. To address these challenges, the research introduces a novel framework titled Person Region Restoration Driven by Perceptual Fidelity (PRRDPF), which combines long-range dependency features with perceptual structure loss for enhanced generative restoration. Specifically, first, the research constructs a high-fidelity distorted person-region dataset via a closed-loop degradation pipeline, addressing the lack of paired datasets. Secondly, a Temporal Gated Fusion (TGF) block is designed to use gated convolutions for selectively recovering high-frequency features while capturing local and global dependencies. Finally, a Structural Similarity Index Measure (SSIM)-based dynamic weighted adversarial loss is proposed to prioritize the restoration of visual texture details. Experimental results validate that PRRDPF significantly outperforms the best models in Peak Signal-to-Noise Ratio (PSNR), SSIM, and Learned Perceptual Image Patch Similarity (LPIPS), effectively mitigating artifacts and enhancing clarity in person visuals. This framework presents a promising approach for intelligent video coding integrated with generative artificial intelligence and holds significant potential for practical applications.
Linyun Liu, Li Yu, Jiaxin Zeng et al.· IEEE Signal Processing Lette...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
