{"results":[{"id":"arxiv_2603.22998","title":"VQ-Jarvis: Retrieval-Augmented Video Restoration Agent with Sharp Vision and Fast Thought","authors":[{"name":"Xuanyu Zhang"},{"name":"Weiqi Li"},{"name":"Qunliang Xing"},{"name":"Jingfen Xie"},{"name":"Bin Chen"},{"name":"Junlin Li"},{"name":"Li Zhang"},{"name":"Jian Zhang"},{"name":"Shijie Zhao"}],"abstract":"Video restoration in real-world scenarios is challenged by heterogeneous degradations, where static architectures and fixed inference pipelines often fail to generalize. Recent agent-based approaches offer dynamic decision making, yet existing video restoration agents remain limited by insufficient quality perception and inefficient search strategies. We propose VQ-Jarvis, a retrieval-augmented, all-in-one intelligent video restoration agent with sharper vision and faster thought. VQ-Jarvis is designed to accurately perceive degradations and subtle differences among paired restoration results, while efficiently discovering optimal restoration trajectories. To enable sharp vision, we construct VSR-Compare, the first large-scale video paired enhancement dataset with 20K comparison pairs covering 7 degradation types, 11 enhancement operators, and diverse content domains. Based on this dataset, we train a multiple operator judge model and a degradation perception model to guide agent decisions. To achieve fast thought, we introduce a hierarchical operator scheduling strategy that adapts to video difficulty: for easy cases, optimal restoration trajectories are retrieved in a one-step manner from a retrieval-augmented generation (RAG) library; for harder cases, a step-by-step greedy search is performed to balance efficiency and accuracy. Extensive experiments demonstrate that VQ-Jarvis consistently outperforms existing methods on complex degraded videos.","source":"arXiv","year":2026,"language":"en","subjects":["cs.CV"],"url":"https://arxiv.org/abs/2603.22998","pdf_url":"https://arxiv.org/pdf/2603.22998","is_open_access":true,"published_at":"2026-03-24T09:40:50Z","score":70},{"id":"arxiv_2603.04032","title":"Multi-Stage Music Source Restoration with BandSplit-RoFormer Separation and HiFi++ GAN","authors":[{"name":"Tobias Morocutti"},{"name":"Emmanouil Karystinaios"},{"name":"Jonathan Greif"},{"name":"Gerhard Widmer"}],"abstract":"Music Source Restoration (MSR) targets recovery of original, unprocessed instrument stems from fully mixed and mastered audio, where production effects and distribution artifacts violate common linear-mixture assumptions. This technical report presents the CP-JKU team's system for the MSR ICASSP Challenge 2025. Our approach decomposes MSR into separation and restoration. First, a single BandSplit-RoFormer separator predicts eight stems plus an auxiliary other stem, and is trained with a three-stage curriculum that progresses from 4-stem warm-start fine-tuning (with LoRA) to 8-stem extension via head expansion. Second, we apply a HiFi++ GAN waveform restorer trained as a generalist and then specialized into eight instrument-specific experts.","source":"arXiv","year":2026,"language":"en","subjects":["cs.SD","cs.LG","eess.AS"],"url":"https://arxiv.org/abs/2603.04032","pdf_url":"https://arxiv.org/pdf/2603.04032","is_open_access":true,"published_at":"2026-03-04T13:10:39Z","score":70},{"id":"arxiv_2603.13089","title":"V-Bridge: Bridging Video Generative Priors to Versatile Few-shot Image Restoration","authors":[{"name":"Shenghe Zheng"},{"name":"Junpeng Jiang"},{"name":"Wenbo Li"}],"abstract":"Large-scale video generative models are trained on vast and diverse visual data, enabling them to internalize rich structural, semantic, and dynamic priors of the visual world. While these models have demonstrated impressive generative capability, their potential as general-purpose visual learners remains largely untapped. In this work, we introduce V-Bridge, a framework that bridges this latent capacity to versatile few-shot image restoration tasks. We reinterpret image restoration not as a static regression problem, but as a progressive generative process, and leverage video models to simulate the gradual refinement from degraded inputs to high-fidelity outputs. Surprisingly, with only 1,000 multi-task training samples (less than 2% of existing restoration methods), pretrained video models can be induced to perform competitive image restoration, achieving multiple tasks with a single model, rivaling specialized architectures designed explicitly for this purpose. Our findings reveal that video generative models implicitly learn powerful and transferable restoration priors that can be activated with only extremely limited data, challenging the traditional boundary between generative modeling and low-level vision, and opening a new design paradigm for foundation models in visual tasks.","source":"arXiv","year":2026,"language":"en","subjects":["cs.CV"],"url":"https://arxiv.org/abs/2603.13089","pdf_url":"https://arxiv.org/pdf/2603.13089","is_open_access":true,"published_at":"2026-03-13T15:39:44Z","score":70},{"id":"arxiv_2511.19314","title":"PRInTS: Reward Modeling for Long-Horizon Information Seeking","authors":[{"name":"Jaewoo Lee"},{"name":"Archiki Prasad"},{"name":"Justin Chih-Yao Chen"},{"name":"Zaid Khan"},{"name":"Elias Stengel-Eskin"},{"name":"Mohit Bansal"}],"abstract":"Information-seeking is a core capability for AI agents, requiring them to gather and reason over tool-generated information across long trajectories. However, such multi-step information-seeking tasks remain challenging for agents backed by language models. While process reward models (PRMs) can guide agents by ranking candidate steps at test-time, existing PRMs, designed for short reasoning with binary judgment, cannot capture richer dimensions of information-seeking steps, such as tool interactions and reasoning over tool outputs, nor handle the rapidly growing context in long-horizon tasks. To address these limitations, we introduce PRInTS, a generative PRM trained with dual capabilities: (1) dense scoring based on the PRM's reasoning across multiple step quality dimensions (e.g., interpretation of tool outputs, tool call informativeness) and (2) trajectory summarization that compresses the growing context while preserving essential information for step evaluation. Extensive evaluations across FRAMES, GAIA (levels 1-3), and WebWalkerQA (easy-hard) benchmarks on multiple models, along with ablations, reveal that best-of-n sampling with PRInTS enhances information-seeking abilities of open-source models as well as specialized agents, matching or surpassing the performance of frontier models with a much smaller backbone agent and outperforming other strong reward modeling baselines.","source":"arXiv","year":2025,"language":"en","subjects":["cs.AI","cs.CL","cs.LG"],"url":"https://arxiv.org/abs/2511.19314","pdf_url":"https://arxiv.org/pdf/2511.19314","is_open_access":true,"published_at":"2025-11-24T17:09:43Z","score":69},{"id":"arxiv_2506.02197","title":"NTIRE 2025 Challenge on RAW Image Restoration and Super-Resolution","authors":[{"name":"Marcos V. Conde"},{"name":"Radu Timofte"},{"name":"Zihao Lu"},{"name":"Xiangyu Kong"},{"name":"Xiaoxia Xing"},{"name":"Fan Wang"},{"name":"Suejin Han"},{"name":"MinKyu Park"},{"name":"Tianyu Zhang"},{"name":"Xin Luo"},{"name":"Yeda Chen"},{"name":"Dong Liu"},{"name":"Li Pang"},{"name":"Yuhang Yang"},{"name":"Hongzhong Wang"},{"name":"Xiangyong Cao"},{"name":"Ruixuan Jiang"},{"name":"Senyan Xu"},{"name":"Siyuan Jiang"},{"name":"Xueyang Fu"},{"name":"Zheng-Jun Zha"},{"name":"Tianyu Hao"},{"name":"Yuhong He"},{"name":"Ruoqi Li"},{"name":"Yueqi Yang"},{"name":"Xiang Yu"},{"name":"Guanlan Hong"},{"name":"Minmin Yi"},{"name":"Yuanjia Chen"},{"name":"Liwen Zhang"},{"name":"Zijie Jin"},{"name":"Cheng Li"},{"name":"Lian Liu"},{"name":"Wei Song"},{"name":"Heng Sun"},{"name":"Yubo Wang"},{"name":"Jinghua Wang"},{"name":"Jiajie Lu"},{"name":"Watchara Ruangsan"}],"abstract":"This paper reviews the NTIRE 2025 RAW Image Restoration and Super-Resolution Challenge, highlighting the proposed solutions and results. New methods for RAW Restoration and Super-Resolution could be essential in modern Image Signal Processing (ISP) pipelines, however, this problem is not as explored as in the RGB domain. The goal of this challenge is two fold, (i) restore RAW images with blur and noise degradations, (ii) upscale RAW Bayer images by 2x, considering unknown noise and blur. In the challenge, a total of 230 participants registered, and 45 submitted results during thee challenge period. This report presents the current state-of-the-art in RAW Restoration.","source":"arXiv","year":2025,"language":"en","subjects":["eess.IV","cs.CV"],"url":"https://arxiv.org/abs/2506.02197","pdf_url":"https://arxiv.org/pdf/2506.02197","is_open_access":true,"published_at":"2025-06-02T19:34:21Z","score":69},{"id":"arxiv_2504.14249","title":"Any Image Restoration via Efficient Spatial-Frequency Degradation Adaptation","authors":[{"name":"Bin Ren"},{"name":"Eduard Zamfir"},{"name":"Zongwei Wu"},{"name":"Yawei Li"},{"name":"Yidi Li"},{"name":"Danda Pani Paudel"},{"name":"Radu Timofte"},{"name":"Ming-Hsuan Yang"},{"name":"Luc Van Gool"},{"name":"Nicu Sebe"}],"abstract":"Restoring multiple degradations efficiently via just one model has become increasingly significant and impactful, especially with the proliferation of mobile devices. Traditional solutions typically involve training dedicated models per degradation, resulting in inefficiency and redundancy. More recent approaches either introduce additional modules to learn visual prompts, significantly increasing the size of the model, or incorporate cross modal transfer from large language models trained on vast datasets, adding complexity to the system architecture. In contrast, our approach, termed AnyIR, takes a unified path that leverages inherent similarity across various degradations to enable both efficient and comprehensive restoration through a joint embedding mechanism, without scaling up the model or relying on large language models. Specifically, we examine the sublatent space of each input, identifying key components and reweighting them first in a gated manner. To unify intrinsic degradation awareness with contextualized attention, we propose a spatial frequency parallel fusion strategy that strengthens spatially informed local global interactions and enriches restoration fidelity from the frequency domain. Comprehensive evaluations across four all-in-one restoration benchmarks demonstrate that AnyIR attains SOTA performance while reducing model parameters by 84% and FLOPs by 80% relative to the baseline. These results highlight the potential of AnyIR as an effective and lightweight solution for further all in one image restoration. Our code is available at: https://github.com/Amazingren/AnyIR.","source":"arXiv","year":2025,"language":"en","subjects":["cs.CV"],"url":"https://arxiv.org/abs/2504.14249","pdf_url":"https://arxiv.org/pdf/2504.14249","is_open_access":true,"published_at":"2025-04-19T09:54:46Z","score":69},{"id":"ss_fd2fadf92b896b824809c27d1b7ceecf5f57ea33","title":"GAN-BASED RECONSTRUCTION OF VINTAGE PRINTS","authors":[{"name":"Mary Praveena J"},{"name":"Vandana Gupta"},{"name":"Smita Rath"},{"name":"Mohit Malik"},{"name":"Sahil Suri"},{"name":"Vishal Ambhore"}],"abstract":"Vintage prints are crucial to preserve the cultural, historical, and artistic heritage and although traditional techniques of restoration are important challenges, physical deterioration, including fading, stains, ripping, and noise are major obstacles to preserve printed images. Manual conservation and classical methods of digital inpainting can be time-consuming, subjective and unable to match the level of fine textuality and stylistic fidelity. This paper presents a GAN-based reconstruction model of the high-quality reconstruction of the damaged vintage prints with the deep generative learning and style-conscious constraints. The suggested method uses an adversarial learning paradigm where a generator network aims at restoring missing structures, textures and tonal continuity and a discriminator network is used to assess realism, stylistic consistency and historical plausibility. The extensive art collection maintained in museums, libraries, and personal collections is filtered, including various patterns of degradation and printing styles. The high-level preprocessing, such as noise normalization, contrast enhancement, degradation-sensitive annotation, and others, facilitates the powerful training. The model considers content similarity preserving loss functions, similarity of perception, and consistency of style as content preserving goals in order to retain artistic integrity. Massive experiments indicate that the suggested structure significantly improves the performance of standard restoration and baseline deep learning structures in terms of structural and perceptual quality and visual authenticity. The effectiveness of the reconstructed outputs as the art historians and painting experts confirm the effectiveness of these measures in preserving original aesthetic character also through qualitative evaluations. The findings in the article suggest that GAN-based reconstruction is a scalable, customizable, and culturally aware way to conserve digital data and allow long-term preservation, accessibility of archival data, and scholarly study of delicate vintage prints.","source":"Semantic Scholar","year":2025,"language":"en","subjects":null,"doi":"10.29121/shodhkosh.v6.i5s.2025.6913","url":"https://www.semanticscholar.org/paper/fd2fadf92b896b824809c27d1b7ceecf5f57ea33","is_open_access":true,"published_at":"","score":69},{"id":"ss_d5e3cd25ada5c2de3625184a07f78cad90823b3c","title":"Implementing 3D Augmented Reality for Increasing Public Outreach of Majapahit Archeological Sites","authors":[{"name":"F. Bioresita"},{"name":"Husnul Hidayat"},{"name":"Arnadi Murtiyoso"},{"name":"E. Moisan"},{"name":"Pierre Grussenmeyer"},{"name":"M. N. Cahyadi"},{"name":"A.B Cahyono"},{"name":"Irham Maulana"}],"abstract":"Abstract. Cultural heritage sites, such as historical buildings and monuments, hold significant artistic, cultural, and historical value which necessitates preservation. This study explores the integration of 3D modelling and Augmented Reality (AR) technologies to digitally document and promote public engagement with heritage sites. Employing both laser scanning and photogrammetry, the research aims to develop accurate and photorealistic 3D models of key archaeological structures from the Majapahit era located in Trowulan, Indonesia—specifically Bajang Ratu, Wringin Lawang, and Brahu. Laser scanning demonstrates superior geometric accuracy and spatial completeness, producing denser point clouds and more detailed meshes, which are essential for precise documentation, restoration planning, and structural analysis. Photogrammetry, while offering lower geometric resolution, excels in capturing high-quality surface textures, making it more suitable for visual representation, public engagement, and AR applications. The findings highlight the benefits of a hybrid approach that combines laser scanning’s spatial precision with photogrammetry’s visual realism. This AR experience, when paired with a printed archaeological site map, allows users to interact with the digital reconstructions, enhancing historical understanding and accessibility. Furthermore, this research contributes to digital conservation efforts and innovative historical learning methods for Indonesian cultural heritage.","source":"Semantic Scholar","year":2025,"language":"en","subjects":null,"doi":"10.5194/isprs-archives-xlviii-m-9-2025-141-2025","url":"https://www.semanticscholar.org/paper/d5e3cd25ada5c2de3625184a07f78cad90823b3c","is_open_access":true,"published_at":"","score":69},{"id":"crossref_10.1186/s12989-024-00592-8","title":"Investigation of pulmonary inflammatory responses following intratracheal instillation of and inhalation exposure to polypropylene microplastics","authors":[{"name":"Taisuke Tomonaga"},{"name":"Hidenori Higashi"},{"name":"Hiroto Izumi"},{"name":"Chinatsu Nishida"},{"name":"Naoki Kawai"},{"name":"Kazuma Sato"},{"name":"Toshiki Morimoto"},{"name":"Yasuyuki Higashi"},{"name":"Kazuhiro Yatera"},{"name":"Yasuo Morimoto"}],"abstract":"Abstract Background Microplastics have been detected in the atmosphere as well as in the ocean, and there is concern about their biological effects in the lungs. We conducted a short-term inhalation exposure and intratracheal instillation using rats to evaluate lung disorders related to microplastics. We conducted an inhalation exposure of polypropylene fine powder at a low concentration of 2 mg/m 3 and a high concentration of 10 mg/m 3 on 8-week-old male Fischer 344 rats for 6 h a day, 5 days a week for 4 weeks. We also conducted an intratracheal instillation of polypropylene at a low dose of 0.2 mg/rat and a high dose of 1.0 mg/rat on 12-week-old male Fischer 344 rats. Rats were dissected from 3 days to 6 months after both exposures, and bronchoalveolar lavage fluid (BALF) and lung tissue were collected to analyze lung inflammation and lung injury. Results Both exposures to polypropylene induced a persistent influx of inflammatory cells and expression of CINC-1, CINC-2, and MPO in BALF from 1 month after exposure. Genetic analysis showed a significant increase in inflammation-related factors for up to 6 months. The low concentration in the inhalation exposure of polypropylene also induced mild lung inflammation. Conclusion These findings suggest that inhaled polypropylene, which is a microplastic, induces persistent lung inflammation and has the potential for lung disorder. Exposure to 2 mg/m 3 induced inflammatory changes and was thought to be the Lowest Observed Adverse Effect Level (LOAEL) for acute effects of polypropylene. However, considering the concentration of microplastics in a real general environment, the risk of environmental hazards to humans may be low.","source":"CrossRef","year":2024,"language":"en","subjects":null,"doi":"10.1186/s12989-024-00592-8","url":"https://doi.org/10.1186/s12989-024-00592-8","pdf_url":"https://link.springer.com/content/pdf/10.1186/s12989-024-00592-8.pdf","is_open_access":true,"citations":31,"published_at":"","score":68.93},{"id":"arxiv_2407.11087","title":"Restore-RWKV: Efficient and Effective Medical Image Restoration with RWKV","authors":[{"name":"Zhiwen Yang"},{"name":"Jiayin Li"},{"name":"Hui Zhang"},{"name":"Dan Zhao"},{"name":"Bingzheng Wei"},{"name":"Yan Xu"}],"abstract":"Transformers have revolutionized medical image restoration, but the quadratic complexity still poses limitations for their application to high-resolution medical images. The recent advent of the Receptance Weighted Key Value (RWKV) model in the natural language processing field has attracted much attention due to its ability to process long sequences efficiently. To leverage its advanced design, we propose Restore-RWKV, the first RWKV-based model for medical image restoration. Since the original RWKV model is designed for 1D sequences, we make two necessary modifications for modeling spatial relations in 2D medical images. First, we present a recurrent WKV (Re-WKV) attention mechanism that captures global dependencies with linear computational complexity. Re-WKV incorporates bidirectional attention as basic for a global receptive field and recurrent attention to effectively model 2D dependencies from various scan directions. Second, we develop an omnidirectional token shift (Omni-Shift) layer that enhances local dependencies by shifting tokens from all directions and across a wide context range. These adaptations make the proposed Restore-RWKV an efficient and effective model for medical image restoration. Even a lightweight variant of Restore-RWKV, with only 1.16 million parameters, achieves comparable or even superior results compared to existing state-of-the-art (SOTA) methods. Extensive experiments demonstrate that the resulting Restore-RWKV achieves SOTA performance across a range of medical image restoration tasks, including PET image synthesis, CT image denoising, MRI image super-resolution, and all-in-one medical image restoration. Code is available at: https://github.com/Yaziwel/Restore-RWKV.","source":"arXiv","year":2024,"language":"en","subjects":["eess.IV","cs.CV"],"url":"https://arxiv.org/abs/2407.11087","pdf_url":"https://arxiv.org/pdf/2407.11087","is_open_access":true,"published_at":"2024-07-14T12:22:05Z","score":68},{"id":"arxiv_2410.08688","title":"Chain-of-Restoration: Multi-Task Image Restoration Models are Zero-Shot Step-by-Step Universal Image Restorers","authors":[{"name":"Jin Cao"},{"name":"Deyu Meng"},{"name":"Xiangyong Cao"}],"abstract":"Despite previous image restoration (IR) methods have often concentrated on isolated degradations, recent research has increasingly focused on addressing composite degradations involving a complex combination of multiple isolated degradations. However, current IR methods for composite degradations require building training data that contain an exponential number of possible degradation combinations, which brings in a significant burden. To alleviate this issue, this paper proposes a new task setting, i.e. Universal Image Restoration (UIR). Specifically, UIR doesn't require training on all the degradation combinations but only on a set of degradation bases and then removing any degradation that these bases can potentially compose in a zero-shot manner. Inspired by the Chain-of-Thought that prompts large language models (LLMs) to address problems step-by-step, we propose Chain-of-Restoration (CoR) mechanism, which instructs models to remove unknown composite degradations step-by-step. By integrating a simple Degradation Discriminator into pre-trained multi-task models, CoR facilitates the process where models remove one degradation basis per step, continuing this process until the image is fully restored from the unknown composite degradation. Extensive experiments show that CoR can significantly improve model performance in removing composite degradations, achieving comparable or better results than those state-of-the-art (SoTA) methods trained on all degradations.","source":"arXiv","year":2024,"language":"en","subjects":["cs.CV","cs.AI"],"url":"https://arxiv.org/abs/2410.08688","pdf_url":"https://arxiv.org/pdf/2410.08688","is_open_access":true,"published_at":"2024-10-11T10:21:42Z","score":68},{"id":"arxiv_2407.13372","title":"Restore Anything Model via Efficient Degradation Adaptation","authors":[{"name":"Bin Ren"},{"name":"Eduard Zamfir"},{"name":"Zongwei Wu"},{"name":"Yawei Li"},{"name":"Yidi Li"},{"name":"Danda Pani Paudel"},{"name":"Radu Timofte"},{"name":"Ming-Hsuan Yang"},{"name":"Nicu Sebe"}],"abstract":"With the proliferation of mobile devices, the need for an efficient model to restore any degraded image has become increasingly significant and impactful. Traditional approaches typically involve training dedicated models for each specific degradation, resulting in inefficiency and redundancy. More recent solutions either introduce additional modules to learn visual prompts significantly increasing model size or incorporate cross-modal transfer from large language models trained on vast datasets, adding complexity to the system architecture. In contrast, our approach, termed RAM, takes a unified path that leverages inherent similarities across various degradations to enable both efficient and comprehensive restoration through a joint embedding mechanism without scaling up the model or relying on large multimodal models. Specifically, we examine the sub-latent space of each input, identifying key components and reweighting them in a gated manner. This intrinsic degradation awareness is further combined with contextualized attention in an X-shaped framework, enhancing local-global interactions. Extensive benchmarking in an all-in-one restoration setting confirms RAM's SOTA performance, reducing model complexity by approximately 82% in trainable parameters and 85% in FLOPs. Our code and models will be publicly available.","source":"arXiv","year":2024,"language":"en","subjects":["cs.CV"],"url":"https://arxiv.org/abs/2407.13372","pdf_url":"https://arxiv.org/pdf/2407.13372","is_open_access":true,"published_at":"2024-07-18T10:26:53Z","score":68},{"id":"arxiv_2406.07435","title":"Beware of Aliases -- Signal Preservation is Crucial for Robust Image Restoration","authors":[{"name":"Shashank Agnihotri"},{"name":"Julia Grabinski"},{"name":"Janis Keuper"},{"name":"Margret Keuper"}],"abstract":"Image restoration networks are usually comprised of an encoder and a decoder, responsible for aggregating image content from noisy, distorted data and to restore clean, undistorted images, respectively. Data aggregation as well as high-resolution image generation both usually come at the risk of involving aliases, i.e.~standard architectures put their ability to reconstruct the model input in jeopardy to reach high PSNR values on validation data. The price to be paid is low model robustness. In this work, we show that simply providing alias-free paths in state-of-the-art reconstruction transformers supports improved model robustness at low costs on the restoration performance. We do so by proposing BOA-Restormer, a transformer-based image restoration model that executes downsampling and upsampling operations partly in the frequency domain to ensure alias-free paths along the entire model while potentially preserving all relevant high-frequency information.","source":"arXiv","year":2024,"language":"en","subjects":["cs.CV","cs.LG","eess.IV"],"url":"https://arxiv.org/abs/2406.07435","pdf_url":"https://arxiv.org/pdf/2406.07435","is_open_access":true,"published_at":"2024-06-11T16:42:17Z","score":68},{"id":"arxiv_2403.18636","title":"A Diffusion-Based Generative Equalizer for Music Restoration","authors":[{"name":"Eloi Moliner"},{"name":"Maija Turunen"},{"name":"Filip Elvander"},{"name":"Vesa Välimäki"}],"abstract":"This paper presents a novel approach to audio restoration, focusing on the enhancement of low-quality music recordings, and in particular historical ones. Building upon a previous algorithm called BABE, or Blind Audio Bandwidth Extension, we introduce BABE-2, which presents a series of improvements. This research broadens the concept of bandwidth extension to \\emph{generative equalization}, a novel task that, to the best of our knowledge, has not been explicitly addressed in previous studies. BABE-2 is built around an optimization algorithm utilizing priors from diffusion models, which are trained or fine-tuned using a curated set of high-quality music tracks. The algorithm simultaneously performs two critical tasks: estimation of the filter degradation magnitude response and hallucination of the restored audio. The proposed method is objectively evaluated on historical piano recordings, showing an enhancement over the prior version. The method yields similarly impressive results in rejuvenating the works of renowned vocalists Enrico Caruso and Nellie Melba. This research represents an advancement in the practical restoration of historical music.","source":"arXiv","year":2024,"language":"en","subjects":["eess.AS","cs.SD"],"url":"https://arxiv.org/abs/2403.18636","pdf_url":"https://arxiv.org/pdf/2403.18636","is_open_access":true,"published_at":"2024-03-27T14:41:39Z","score":68},{"id":"arxiv_2404.02606","title":"Chiral symmetry restoration at finite temperature in a model with manifest confinement","authors":[{"name":"L. Ya. Glozman"},{"name":"A. V. Nefediev"},{"name":"R. Wagenbrunn"}],"abstract":"Multiple lattice evidences support the existence of a confining but chirally symmetric regime of QCD above the chiral symmetry restoration crossover at Tch ~ 155 MeV. This regime is characterised by an approximate chiral spin symmetry of the partition function, which is a symmetry of the colour charge and the confining electric part of the QCD Lagrangian. It is traditionally believed that confinement should automatically induce spontaneous breaking of chiral symmetry, which would preclude the existence of a confining but chirally symmetric regime of QCD at high temperatures. We employ a well-known solvable quark model for QCD in 3+1 dimensions that is chirally symmetric and manifestly confining and argue that while confinement indeed induces dynamical breaking of chiral symmetry at T=0, a chiral restoration phase transition takes place at some critical temperature Tch. Above this temperature, the spectrum of the model consists of chirally symmetric hadrons with approximate chiral spin symmetry.","source":"arXiv","year":2024,"language":"en","subjects":["hep-ph","hep-lat","hep-th","nucl-th"],"url":"https://arxiv.org/abs/2404.02606","pdf_url":"https://arxiv.org/pdf/2404.02606","is_open_access":true,"published_at":"2024-04-03T09:52:57Z","score":68},{"id":"arxiv_2407.00261","title":"Generative Iris Prior Embedded Transformer for Iris Restoration","authors":[{"name":"Yubo Huang"},{"name":"Jia Wang"},{"name":"Peipei Li"},{"name":"Liuyu Xiang"},{"name":"Peigang Li"},{"name":"Zhaofeng He"}],"abstract":"Iris restoration from complexly degraded iris images, aiming to improve iris recognition performance, is a challenging problem. Due to the complex degradation, directly training a convolutional neural network (CNN) without prior cannot yield satisfactory results. In this work, we propose a generative iris prior embedded Transformer model (Gformer), in which we build a hierarchical encoder-decoder network employing Transformer block and generative iris prior. First, we tame Transformer blocks to model long-range dependencies in target images. Second, we pretrain an iris generative adversarial network (GAN) to obtain the rich iris prior, and incorporate it into the iris restoration process with our iris feature modulator. Our experiments demonstrate that the proposed Gformer outperforms state-of-the-art methods. Besides, iris recognition performance has been significantly improved after applying Gformer.","source":"arXiv","year":2024,"language":"en","subjects":["eess.IV","cs.CV"],"doi":"10.1109/ICME55011.2023.00094","url":"https://arxiv.org/abs/2407.00261","pdf_url":"https://arxiv.org/pdf/2407.00261","is_open_access":true,"published_at":"2024-06-28T23:20:57Z","score":68},{"id":"ss_167b2827cdc57140fd3d600b6fec26241697a25c","title":"3D PRINTING TECHNOLOGY AND ITS APPLICATION IN THE CONSERVATION AND RESTORATION OF PORCELAIN","authors":[{"name":"Y. Zhou"},{"name":"E. Aura-Castro"},{"name":"E. Nebot Díaz"}],"abstract":"Abstract. This paper describes the Charter on the preservation of the digital heritage and it also presents the five main printing technologies, FDM (Fused Deposition Modeling), SLA (Stereo Lithography Apparatus), DLP (Digital Light Processing), LCD (Liquid Crystal Display) and SLS (selective laser sintering) technology. Besides, It presents a methodology to process the restoration of a porcelain bowl through three-dimensional scanning and FDM (Fused Deposition Modeling) and LCD (Liquid Crystal Display) 3D printing technologies. First, the porcelain bowl is scanned with the scanner. Then the modeling software is used to reconstruct the missing part that the bowl presents. Finally, the reconstructed 3D fragment model is printed, completed and chromatically reintegrated. In order to optimize this restoration method to achieve the best visual effect similar to the appearance of the porcelain bowl, different printing materials are used for testing. This type of restoration method can improve the final appearance of the intervened area, minimize the operation of the object and can also make it be shaped quickly and precisely.","source":"Semantic Scholar","year":2023,"language":"en","subjects":null,"doi":"10.5194/isprs-annals-x-m-1-2023-301-2023","url":"https://www.semanticscholar.org/paper/167b2827cdc57140fd3d600b6fec26241697a25c","pdf_url":"https://isprs-annals.copernicus.org/articles/X-M-1-2023/301/2023/isprs-annals-X-M-1-2023-301-2023.pdf","is_open_access":true,"citations":3,"published_at":"","score":67.09},{"id":"arxiv_2309.15490","title":"Survey on Deep Face Restoration: From Non-blind to Blind and Beyond","authors":[{"name":"Wenjie Li"},{"name":"Mei Wang"},{"name":"Kai Zhang"},{"name":"Juncheng Li"},{"name":"Xiaoming Li"},{"name":"Yuhang Zhang"},{"name":"Guangwei Gao"},{"name":"Weihong Deng"},{"name":"Chia-Wen Lin"}],"abstract":"Face restoration (FR) is a specialized field within image restoration that aims to recover low-quality (LQ) face images into high-quality (HQ) face images. Recent advances in deep learning technology have led to significant progress in FR methods. In this paper, we begin by examining the prevalent factors responsible for real-world LQ images and introduce degradation techniques used to synthesize LQ images. We also discuss notable benchmarks commonly utilized in the field. Next, we categorize FR methods based on different tasks and explain their evolution over time. Furthermore, we explore the various facial priors commonly utilized in the restoration process and discuss strategies to enhance their effectiveness. In the experimental section, we thoroughly evaluate the performance of state-of-the-art FR methods across various tasks using a unified benchmark. We analyze their performance from different perspectives. Finally, we discuss the challenges faced in the field of FR and propose potential directions for future advancements. The open-source repository corresponding to this work can be found at https:// github.com/ 24wenjie-li/ Awesome-Face-Restoration.","source":"arXiv","year":2023,"language":"en","subjects":["cs.CV"],"url":"https://arxiv.org/abs/2309.15490","pdf_url":"https://arxiv.org/pdf/2309.15490","is_open_access":true,"published_at":"2023-09-27T08:39:03Z","score":67},{"id":"crossref_10.61882/kcr.6.3.52","title":"Conservation and Restoration Report of Sassanian Jar from Varamin Plain, Iran (An Overview of the Rules and Principles of Conservation and Restoration of Pottery Objects.)","authors":[{"name":"Fatemeh Alimirzaei"}],"abstract":"","source":"CrossRef","year":2023,"language":"en","subjects":null,"doi":"10.61882/kcr.6.3.52","url":"https://doi.org/10.61882/kcr.6.3.52","pdf_url":"https://kcr.richt.ir/article-1-94-en.pdf","is_open_access":true,"published_at":"","score":67},{"id":"ss_4c6320f5a60b4d885ebb98148957bdcdc0e67afa","title":"INVESTIGATION AND CONSERVATION OF A PRIVATE PHOTOGRAPHIC COLLECTION OF ALBUMEN PRINTS, EGYPT","authors":[{"name":"E. H."},{"name":"A. M."},{"name":"M. M."}],"abstract":": Albumen prints are the most important photographic prints of the late 19 th century. It is basically composed of two layers : the first layer is the paper support (i.e. cellulose), and the second layer is the image layer (i.e. image silver particles embedded in an albumen binder layer). There are several factors threatening the permanence of albumen prints (e.g., fluctuating temperatures and relative humidity, frequent han -dling, air pollution, light, and improper storage and display). Unlike other paper objects, photographs have special cons ervation requirements due to their complex and unique nature . A private collection was selected for this study. The collection consists of three albumen prints from Francis Amen’s photo collection, which originally belonged to the Elhagar family. Francis Amin is a photo Egypt. The prints back to 1890. The photographs were characterised and studied by visual inspection, digital microscopy, Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), and energy dispersive X-ray spectroscopy (EDX). Microbiological studies were carried out in the microbiology laboratory at the Faculty of Archaeology at Cairo University. Results revealed that the albumen layer suffers from cracks and chemical degradation, and the secondary supports suffer from both oxidation and hydrolysis. Based on the obtained results, the following conservation procedures were selected and carried out: disinfection, dry cleaning, tear mending and compensating for losses, remounting, retouching, and rehousing.","source":"Semantic Scholar","year":2022,"language":"en","subjects":null,"doi":"10.21608/ejars.2022.246575","url":"https://www.semanticscholar.org/paper/4c6320f5a60b4d885ebb98148957bdcdc0e67afa","pdf_url":"https://doi.org/10.21608/ejars.2022.246575","is_open_access":true,"citations":3,"published_at":"","score":66.09}],"total":490964,"page":1,"page_size":20,"sources":["DOAJ","arXiv","CrossRef","Semantic Scholar"],"query":"Conservation and restoration of prints"}