paperlist.ai

AI conference papers

AI conference paper library

Search papers from 30+ top conferences, including NeurIPS, ICLR, ICML, CVPR, ACL and AAAI. Filter by keyword, then open any paper for its abstract, authors and links.

arXiv

Mean-square and sublinear convergence of a stochastic proximal point algorithm in metric spaces of nonpositive curvature

Nicholas Pischke

We define a stochastic variant of the proximal point algorithm in the general setting of nonlinear Hadamard spaces for approximating zeros of the mean of a stochastically perturbed monotone vector field. Generalizing previous work by P. Bianchi, we prove the convergence of this method under a suitable strong monotonicity assumption in (separable) Hilbert-Hadamard spaces, that is assuming that all tangent spaces isometrically embed into Hilbert spaces (covering, but not being limited to, the setting of Hadamard manifolds). Moreover, our convergence proof is fully effective and allows for the construction of explicit rates of convergence for the iteration towards the (unique) solution both in mean and almost surely. These rates are moreover highly uniform, being independent of most data surrounding the iteration, space or distribution. In that generality, these rates are novel already in the context of Hilbert spaces. Sublinear nonasymptotic guarantees under additional second-moment conditions on the Yosida approximates and special cases of stochastic convex minimization are discussed.

arXiv

Stability of Transformers under Layer Normalization

Kelvin Kan, Xingjian Li, Benjamin J. Zhang, Tuhin Sahai, Stanley Osher, Krishna Kumar, Markos A. Katsoulakis

Despite their widespread use, training deep Transformers can be unstable. Layer normalization, a standard component, improves training stability, but its placement has often been ad-hoc. In this paper, we conduct a principled study on the forward (hidden states) and backward (gradient) stability of Transformers under different layer normalization placements. Our theory provides key insights into the training dynamics: whether training drives Transformers toward regular solutions or pathological behaviors. For forward stability, we derive explicit bounds on the growth of hidden states in trained Transformers. For backward stability, we analyze how layer normalization affects the backpropagation of gradients, thereby explaining the training dynamics of each layer normalization placement. Our analysis also guides the scaling of residual steps in Transformer blocks, where appropriate choices can further improve stability and performance. Our numerical results corroborate our theoretical findings. Beyond these results, our framework provides a principled way to sanity-check the stability of Transformers under new architectural modifications, offering guidance for future designs.

arXiv

Optimization as a Dynamical System: Generative Schedules from Latent ODEs

Matt L. Sampson, Peter Melchior

We present a new meta-learning method to determine the optimal learning rate schedule for gradient descent. It leverages training runs from a hyperparameter search to learn a latent representation of the training process, which is modeled as a dynamical system. Given current training metrics, it predicts the future learning rate schedule with the best long-term validation performance. Our scheduler generalizes beyond previously observed training dynamics and creates specialized schedules that deviate noticeably from even the best-performing parametric functions. It outperforms all baselines we compare to on results for image classification with CNN and ResNet models as well as for next-token prediction with a transformer model. The trained models are located in flatter regions of the loss landscape and thus provide better generalization than those trained with other schedules. Our method is computationally efficient, optimizer-agnostic, and can easily be layered on top of ML experiment-tracking platforms to streamline training of neural networks from scratch.

arXiv

Robot guide with multi-agent control and automatic scenario generation with LLM

Elizaveta D. Moskovskaya, Anton D. Moscowsky

The article describes the development of a hybrid social robot control architecture to overcome the limitations of traditional approaches, where behavior scripts manually synchronize the robot's actions and text, and existing methods focus primarily on short dialogue responses. The architecture of the proposed system combines a multi-agent resource management system with automatic generation of behavior scenarios based on large language models. This system automates the preparation of text and commands for the robot's non-verbal behavior for extended narratives and resolves resource conflicts between multiple execution mechanisms. The system was tested on the MENTOR-1 tour guide robot, for which it successfully generated scenarios automatically and demonstrated more natural and rich behavior compared to existing approaches. The proposed approach provides full automation of both scenario preparation and execution through efficient resource management, enhancing the quality of social robot interaction in long-term storytelling tasks.

arXiv

Optimization-based Online Conformal Prediction for Multi-step Forecasting

Ruipu Li, Daniel Menacho, Alexander Rodríguez

Conformal prediction (CP) provides distribution-free coverage guarantees, making it well suited for uncertainty quantification in time series forecasting. However, existing methods often struggle with multi-step settings: they either calibrate horizons independently---ignoring temporal correlations---or enforce strict simultaneous coverage, resulting in overly conservative intervals. In this work, we propose O$^2$CP: Optimization-Based Online Conformal Prediction, a framework that augments a broad family of online CP methods with cross-horizon optimization while preserving their long-term coverage guarantees. We first characterize this family of methods, showing that long-term coverage is preserved as long as, at each forecast horizon, the selected control variable remains within an admissible set around the method's nominal output. Building on this result, O$^2$CP uses a two-layer design: the first layer constructs these admissible sets from the underlying online CP updates, and the second performs constrained optimization across horizons within them, jointly modeling the cross-horizon distributions to minimize a user-specified objective. Extensive experiments on real-world datasets---including autonomous driving, climate forecasting, and public health---demonstrate that O$^2$CP consistently outperforms state-of-the-art baselines, achieving target coverage with significantly sharper prediction intervals and reduced regret over long horizons.

arXiv

Pay Attention to Attention Distribution: A New Local Lipschitz Bound for Transformers

Nikolay Yudin, Sergei Kudriashov, Alexander Gaponov, Maxim Rakhuba

We introduce a novel upper bound on the local Lipschitz constant of the dot-product self-attention block showing its dependence on the attention map distributions. The proposed bound is not only tighter than the prior art, but for the first time, reveals how the distribution of attention probabilities shapes the local Lipschitz constant of the self-attention block. The theoretical basis of the proposed upper bound lies in the refined closed-form upper bounds on singular values of the Jacobian of softmax function. Leveraging these theoretical insights, we introduce JaSMin (Jacobian Softmax norm Minimization), a lightweight regularizer that directly controls the local Lipschitz constant of each block and, consequently, the entire model. Additionally, we discuss how the nature of the attention map distribution contributes to the gradient dynamics and, consequently, transformer training stability.

arXiv

Minimal Ingredients for Reward Assignment from Expert Demonstrations

Zixuan Dong, Yumi Omori, Keith Ross

Reward assignment from scarce demonstrations is a key challenge in both offline and online imitation learning. A common and intuitive strategy assigns rewards according to how closely learner trajectories match expert demonstrations. Although this principle underlies many existing methods, the core ingredients that drive performance remain systematically underexplored. We therefore ask: what is the minimal structure that reward assignment must encode to achieve effective downstream RL performance across settings? We approach this question along two design axes: proximity approximation and temporal alignment. Across 32 benchmarks spanning offline and online settings, and with three downstream RL algorithms, our empirical findings suggest: (1) In offline regimes, proximity alone captures the reward structure necessary for effective offline RL, while (2) lightweight temporal correspondence provides consistent gains that are modest offline but essential online or in the presence of multiple demonstrations. We further complement our offline results with a lightweight theory characterizing when simple proximity approximation suffices. Overall, these findings advocate algorithmic minimalism in reward design before introducing complex schemes in both offline and online imitation learning.

arXiv

CHIME: A Case for Efficient Long-Context Attention-FC Disaggregated Inference with DIMM-PIM

Qingyuan Liu, Liyan Chen, Haocheng Wang, Yanning Yang, Dong Du, Zhigang Mao, Naifeng Jing, Yubin Xia, Haibo Chen

Attention-FC Disaggregated (AFD) LLM inference systems offload memory-bound Attention operations to memory-rich accelerators (e.g., CPUs, HBM-PIM) while retaining compute-bound Fully-Connected (FC) operations on GPUs. In this paper, we first design a Disaggregated Roofline Model (DRM) to characterize AFD performance, revealing that system throughput is constrained by the accelerator's limiting factor: either memory bandwidth or capacity. We observe that prior AFD systems often overlook these constraints and fail to balance them, leading to resource underutilization or constrained throughput. Therefore, we propose CHIME, the first AFD system integrating DIMM-PIM, which is a case of the new accelerator that strikes the balance with scalable capacity and bandwidth. To address the synchronization challenges inherent to the distributed cooperating DRAM chips in DIMM-PIM, CHIME employs bubble-free pipelining and hybrid-grained re-layout for efficient attention computation. Furthermore, it maximizes cross-device resource utilization via rankset-granular communication-computation overlapping and alignment-predicting scheduling. Evaluations show CHIME achieves up to 5.15$\times$ speedup over state-of-the-art HBM-PIM solutions.

arXiv

Decentralized Indoor Localization Based on A Sparse Gaussian Process with Reduced-Dimensional Inputs for Real-Time Sensing and Training on IoT Devices

Zhe Tang, Sihao Li, Zichen Huang, Guandong Yang, Kyeong Soo Kim, Jeremy S. Smith, Zhaowei Zhu, Qi Xuan

As a large number of Internet of Things (IoT) devices are deployed in the field, there arises huge potential of edge computing for indoor localization on those devices. Conventional indoor localization based on a centralized server with substantial computational resources, often covering a number of multistory buildings, cannot easily adapt to time-varying indoor electromagnetic environments due to its high cost of fingerprint database update and model retraining; the centralized server is also susceptible to security breaches. To address these issues, we propose a decentralized indoor localization framework, leveraging models based on a Sparse Gaussian Process with Reduced-dimensional Inputs (SGP-RI) deployed to IoT devices for a smaller service area, which can quickly adapt to time-varying indoor electromagnetic environments through real-time sensing and retraining. The experimental results based on a multibuilding, multifloor static database and a single-building, single-floor dynamic database, demonstrate the feasibility of the proposed framework, where the SGP-RI with less than half the training samples can produce localization performance comparable to the standard Gaussian process (GP) with the whole training samples.

arXiv

Quantum Generative Diffusion Model: A Fully Quantum-Mechanical Model for Generating Quantum State Ensemble

Chuangtao Chen, Qinglin Zhao, Mengchu Zhou, Zhimin He, Zhili Sun, Haozhen Situ

Mixed quantum states are the native description of many physically important quantum systems, making their generation a fundamental task in quantum information processing. However, constructing a diffusion process that generates density operators while keeping every reverse step physically valid remains nontrivial. This work introduces Quantum Generative Diffusion Model (QGDM), a fully quantum-mechanical model whose forward and backward processes are grounded in quantum channel theory. Through a non-unitary forward process, any target quantum state can be transformed into a completely mixed state. A trainable backward process recovers the former from the latter. We introduce partial trace to make the backward process non-unitary, and share parameters across timesteps by incorporating temporal information as an input. We present QGDM's resource-efficient version to reduce auxiliary qubits while preserving generative capabilities. We theoretically analyze the denoising design, showing it avoids a low-loss shortcut that traps training and cause generation failure. Simulations confirm that QGDM outperforms quantum generative adversarial networks on random pure- and mixed-state generation, with better noise robustness than other quantum generative models and a task-specialized approach for practical Gibbs state generation. Hence, QGDM provides a channel-based diffusion framework for learning fixed mixed-state targets, extending quantum generative modeling toward realistic quantum information settings.

arXiv

Generative Gap Filling

Yonathan A. Arbel, David A. Hoffman

Most contract litigation turns on contracts that imperfectly record parties' bargains. When the parties' dispute can't be solved by interpreting the text, courts fill the gap. Scholars have long assumed that the remaining text runs out quickly, and provides thin evidence of the actual deal on the disputed point. On that view, a judge who supplies the missing term must be drawing on something else, from commercial defaults to her own policy preferences. Despite generations of work, courts have no real alternative to such unruly methods. We tested that assumption. Taking real contracts, we masked a term the parties had negotiated and asked readers to predict what we removed. Lay respondents recovered the hidden term about half the time, twice what chance predicts. Law students and lawyers did marginally better. But large language models, given nothing but the rest of the contract, recovered it nearly nine times in ten. The deal, in short, testifies to far more of the agreement than the literature assumes, including terms the parties never wrote. A contract, we argue, is like a radio signal from far away. Even when incomplete, enough of the message is carried elsewhere that the missing part can be reconstructed with the right receiver. True gaps are rarer than supposed. Courts can weigh model predictions as ordinary, contestable evidence, and parties can discipline the practice with "Choice of Model" clauses.

arXiv

Glance, Scrutinize, and Think: Advancing Video Anomaly Detection from Training-Free to Agentic Reasoning

Shibo Gao, Peipei Yang, Xu-Yao Zhang, Linlin Huang

Video Anomaly Detection (VAD) aims to identify anomalous events and localize their temporal intervals. Existing approaches exhibit a "when-what" dissociation: traditional DNN-based methods localize when anomalies occur but lack semantic understanding, whereas LLM-based methods explain what happens but neglect precise temporal grounding. We attribute this to the absence of a unified reasoning paradigm. Inspired by how humans inspect surveillance videos - glancing globally to form temporal hypotheses, scrutinizing suspicious segments, and thinking iteratively to correct errors - we study this global-to-local paradigm from two perspectives. We first propose Glance then Scrutinize (GtS), a training-free framework using static and dynamic textual guidance for coarse-to-fine anomaly grounding and understanding, balancing accuracy and speed. To break the ceiling imposed by frozen external modules, we further propose a tool-augmented agentic VAD method, where a multimodal large language model learns to invoke a video cropping tool, inspect densely resampled frames, and self-correct mislocalized hypotheses, via cold-start supervised fine-tuning followed by reinforcement learning with a joint answer-grounding reward. For training and evaluation, we extend our prior VAGU benchmark into VAGU-T (Video Anomaly Grounding, Understanding, and Thinking), comprising 7,567 real-world videos over 21 anomaly categories with human-validated grounding, explanations, QA pairs, and chain-of-thought tool-calling traces. We further introduce JeAUG, a metric jointly evaluating semantic interpretability and temporal precision. Experiments show that GtS substantially surpasses training-free baselines, while the agentic model delivers both higher accuracy and faster inference.

arXiv

ReCBM: Uncertainty-Gated Relational Reasoning for Concept Bottleneck Models

An Sui, Yuzhu Li, Fuping Wu, Xiahai Zhuang

Concept Bottleneck Models (CBMs) provide an interpretable framework by grounding predictions in human-understandable concepts, enabling semantic inspection and test-time intervention. Recent variants have improved CBMs through richer concept representations, uncertainty estimation, and dependency modeling. However, robust reasoning under unreliable concept states remains underexplored. Without such reasoning, misleading semantic evidence can propagate through the bottleneck, compromising both explanations and downstream predictions. To address this issue, we propose ReCBM, an uncertainty-gated relational reasoning framework for CBMs. ReCBM introduces semantically defined concept relations into the bottleneck and uses uncertainty to guide their refinement. By modeling co-occurrence, implication, and exclusion, ReCBM specifies how evidence is exchanged across concepts, while uncertainty modulates the contribution of each concept during this process. Experiments across diverse datasets showed that ReCBM improved concept and task recovery under missing and flipped concepts, supported uncertainty-aware intervention, and extracted compact task-relevant concept subsets without degrading downstream performance.

arXiv

LoRCA: LoRA Cycle Adaptation for Histology to HiP-CT Translation with DINOv3

Yang Zhou, Edoardo Occhipinti, Banboye Kidzeru Elvis, Jishizhan Chen, Stathis Megas, Joseph Brunet, Joanna Purzycka, Theresa Urban, Hector Dejea, Sarah Amalia Teichmann, Menna R Clatworthy, Paul Tafforeau, Peter D Lee, Claire L Walsh

Hierarchical Phase-Contrast Tomography (HiP-CT) is a synchrotron based X-ray imaging technique that enables non-destructive, volumetric imaging of intact organs with multi-resolutions bridging 20 $μm$/voxel for whole organs to near-cellular resolution ($\sim$0.8 $μm$/voxel) in local regions. This offers the opportunity to bring volumetric whole-organ context to histology. However, nonlinear registration between H\&E histology and HiP-CT volumes is challenging due to the differences in feature representations of different colour spaces. Synthesis-before-registration methods have shown strong results in histology-to-MRI and histology-to-CT alignment. However, existing approaches either rely on manual anatomical contours or are trained from scratch without semantic constraints, limiting their generalisability to soft tissue organs and novel modalities. We propose LoRCA (LoRA Cycle Adaptation), a cycle consistent style translation framework built on a shared frozen DINOv3 with modality-specific LoRA adapters, learning modality-specific representations that are decoded and adversarially trained. LoRCA enables structure-preserving translation without requiring paired training data. The frozen backbone is intended to be a structural anchor that prevents content drift by preserving pretrained semantic-extraction capability. We evaluate translation quality using Fréchet Inception Distance (FID) and structural fidelity via mutual information and Canny edge preservation. LoRCA outperforms CycleGAN in both translation quality and structural consistency. As a preliminary indicator of downstream registration utility, we find that style-translated images yield increased feature correspondences under MatchAnything on manually aligned HiP-CT and histology test pairs, suggesting that LoRCA-style translation is a promising step towards 2D histological sections to 3D HiP-CT volumes registration.

arXiv

Implicit representations are dead. Long live explicit primitives!

Nil Stolt-Ansó, Maik Dannecker, Wenqi Huang, Andras Jakab, Daniel Rueckert

Continuous parameterization of medical data has emerged as a powerful paradigm for resolution-independent image representation. While Implicit Neural Representations offer high fidelity and compact storage, their reliance on global Multi-Layer Perceptrons incurs sizeable computational costs, large memory requirements, and extensive optimization times. As medical imaging trends towards ever-more detailed, high-resolution volumes, these costs impose significant bottlenecks in the applicability of implicit approaches. Recently, explicit Gaussian-based primitives have revolutionized the representation learning paradigm by trading deep network evaluations for localized, rasterization-friendly primitives. In this paper, we present a comprehensive, cross-dimensional evaluation of Gaussian representations against implicit approaches for medical imaging applications. First, we outline a theoretical overview on the mathematical properties offered by explicit primitives beyond what is capable under the implicit neural paradigm. Subsequently, we benchmark the computational performance on two demanding image datasets: 2D microscopy histology and 3D lung Computed Tomography (CT). Our experiments demonstrate that Gaussian representations consistently match or surpass reconstruction metrics compared to implicit methods across all compression factors, while displaying significantly lower optimization times, and memory requirements. Together with the compelling mathematical properties offered by explicit primitives, these findings motivate the wider adoption of Gaussian representations and position them as an attractive direction for future research in medical imaging.

arXiv

Pre- to Post-Contrast Synthesis of Breast DCE-MRI using Latent Bridge Matching

Sina Amirrajab, Zohaib Salahuddin, Henry C Woodruff, Philippe Lambin

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is central to breast cancer imaging, but gadolinium administration increases scan burden and motivates contrast-reduced alternatives, including synthetic contrast generation. We propose a latent bridge matching (LBM) framework for synthesizing peak-enhanced breast DCE-MRI from pre-contrast images in the MAMA-SYNTH challenge setting. Instead of starting from Gaussian noise as in conventional latent diffusion models (LDMs), the proposed model learns a conditional bridge between paired pre-contrast and peak-enhanced VAE latents. A latent UNet predicts the remaining correction from intermediate bridge states to the peak-enhanced latent, enabling iterative refinement while keeping the trajectory anchored to patient-specific anatomy. We evaluated two LBM conditioning variants on 91 DUKE validation cases. For the tumor-conditioned variant, tumor masks were used as conditioning inputs. Tumor-conditioning improved performance compared with pre-contrast conditioning, reducing MSE from 1.023 to 0.940 and FRD from 7.523 to 4.716, while increasing tumor SSIM from 0.355 to 0.429. The tumor-conditioned LBM also outperformed the evaluated LDM baseline on this validation cohort. These results suggest that latent bridge matching is a promising pre-contrast-anchored formulation for virtual contrast enhancement, while further work is needed to validate generalization and remove dependence on ground-truth tumor masks at inference.

arXiv

Robustness of transferability estimation metrics for medical imaging

Niclas Claßen, Théo Sourget, Dovile Juodelyte, Rob van der Goot, Veronika Cheplygina

In transfer learning, the choice of source model largely influences the performance on a target dataset. Still, selecting a fitting source remains a challenging task, especially in medical imaging where one has to decide between models pre-trained on off-the-shelf options, such as ImageNet, and domain specific datasets. Transferability estimation (TE) metrics address this problem by aiming to predict the best performing source model in a computationally cost effective way. However, previous work has reported conflicting TE metric performances due to differences in experimental setups. Moreover, most TE metrics are designed for and evaluated on natural images, while being optimized for accuracy, whereas in medical imaging metrics that are more robust to class imbalance are typically used. We study the impact of varying the target dataset as an isolated factor, by constructing miniature populations of different sample sizes and random seeds. In addition, we investigate the influence of the evaluation metric used to obtain the reference ranking. We find that small modifications to the target dataset change the rankings. Furthermore, we show that the choice of evaluation metric affects the reference rankings and therefore the evaluation of TE metrics. Overall, we observe a low agreement between rankings from TE metrics and reference. The code, model checkpoints and data splits used in this work are available through https://github.com/niclasclassen/robustness-of-transferability-estimation-metrics-for-medical-imaging.

arXiv

Transformer Geometry Observatory TGO-IV: Developmental Topology Observatory

Kaustubh Kapil, Kishor P. Upla

Transformers have had a profound impact on the world of language processing and computer vision. As efforts to answer the million-dollar question of ``How does a Transformer learn?" have been increasing, existing interpretability studies primarily analyze representations at isolated layers or the network as a whole, while the developmental evolution of individual representations and its manifolds across transformer layers remains underexplored. With this work, we aim at providing a comprehensive analysis of the evolution of representations as the representation point cloud transforms across the layers; thereby attempting to isolate layers or establish a trend which comes closer to justifying how and when raw input representations evolve into task-relevant feature representations. Thus, Transformer Geometry Observatory-TGO-IV introduces a topological framework for analysing the evolution of Transformer representations through the lens of Persistent Homology. Rather than studying local geometric properties alone, TGO-IV constructs Vietoris--Rips simplicial complexes from token-level representation point clouds and investigates the evolution of their persistent topological signatures across Transformer layers. The proposed framework comprises complementary topological observatories including Persistence Diagrams, Barcode Diagrams, Betti Curves, Persistence Landscapes, Bottleneck Distance, and Wasserstein Distance, enabling a comprehensive analysis of how the global topology of representation point clouds develops throughout the forward pass.

arXiv

Energy and Performance Benchmarking of Deep Learning Models for Breast Cancer Detection

Samar Garrab, Ghada Achour

Recent advances in machine learning have greatly improved breast cancer detection, enabling more accurate and timely diagnosis. Deep learning (DL) models show strong potential for medical image analysis; however, as their architectural complexity increases, their environmental impacts are becoming a growing concern. In this paper, we present a comparative analysis of seven DL models for breast cancer detection on two medical datasets: Breast Ultrasound and BreakHis 400X. The evaluated architectures range from Convolutional Neural Networks (CNNs) and transformers to hybrid models. In addition to performance metrics, we assess CO2 emissions during both training and inference. Our results show that EfficientNet and ResNet consistently deliver strong performance, although with higher CO2 emissions. The selected transformers, such as DeiT-Tiny, perform competitively on both datasets, whereas DenseNet121 achieves lower accuracy. On the Breast Ultrasound Dataset, DeiT provides the most favourable balance between accuracy and energy consumption, whereas on the BreakHis dataset, the ViT and Swin models achieve the best results. Overall, our findings indicate that no single architecture category from the evaluated ones consistently dominates across the two selected datasets. Our results highlight the importance of jointly considering performance, emissions, and dataset characteristics when selecting models for medical applications.

arXiv

Structural Guidance for Unified Joint Demosaicing and Denoising

Qixin Zheng, Ping Chen, Qiangqiang Shen, Haijin Zeng

Joint demosaicing and denoising is a fundamental step in camera image signal processing, yet remains challenging because different Bayer-like color filter arrays (CFAs) and sensor noise jointly corrupt both color sampling and image content. Existing unified restoration networks explicitly model CFA geometry but are still driven primarily by pixel-level supervision, making them prone to structural degradation around edges, repetitive textures, and moiré patterns where local evidence is unreliable. We attribute this limitation partly to the absence of explicit structural guidance beyond pixel-level reconstruction supervision. Motivated by this observation, we propose a structural-guided unified restoration framework that injects pretrained structural knowledge into CFA-aware image restoration. Our model receives a unified five-channel observation consisting of the raw mosaic, CFA masks, and a noise-level map. A SwinIR restoration branch reconstructs pixel details under CFA-conditioned modulation, while a parallel structural reasoning branch extracts complementary structural cues from a sparse pseudo-RGB observation. To bridge the substantial domain gap between sparse noisy sensor data and the natural-image pretraining domain of the structural encoder, we introduce a lightweight trainable adapter before residually fusing structural and restoration features. A shared decoder jointly predicts the restored RGB image and an auxiliary clean mosaic, providing supervision in both image and sensor domains. Extensive experiments across multiple CFA patterns and noise levels demonstrate consistent improvements over state-of-the-art unified and CFA-specific methods, indicating that adapted structural priors can enhance robust camera image restoration. The source codes and dataset are provided in the supplementary material.

Keep browsing

More papers from top AI conferences

30+ top conferences, by conference, year, and topic.