Sparsity-Aware Optimal Transport for Unsupervised Restoration Learning

• URL: http://arxiv.org/abs/2305.00273v1
• Date: Sat, 29 Apr 2023 15:09:48 GMT
• Title: Sparsity-Aware Optimal Transport for Unsupervised Restoration Learning
• Abstract summary: In this paper, we exploit the sparsity of degradation in the unsupervised restoration learning framework to significantly boost its performance on complex restoration tasks. Experiments on real-world super-resolution, deraining, and dehazing demonstrate that SOT can improve the PSNR of OT by about 2.6 dB, 2.7 dB and 1.3 dB, respectively.
• Score: 17.098664719423404
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent studies show that, without any prior model, the unsupervised restoration learning problem can be optimally formulated as an optimal transport (OT) problem, which has shown promising performance on denoising tasks to approach the performance of supervised methods. However, it still significantly lags behind state-of-the-art supervised methods on complex restoration tasks such as super-resolution, deraining, and dehazing. In this paper, we exploit the sparsity of degradation in the OT framework to significantly boost its performance on these tasks. First, we disclose an observation that the degradation in these tasks is quite sparse in the frequency domain, and then propose a sparsity-aware optimal transport (SOT) criterion for unsupervised restoration learning. Further, we provide an analytic example to illustrate that exploiting the sparsity helps to reduce the ambiguity in finding an inverse map for restoration. Experiments on real-world super-resolution, deraining, and dehazing demonstrate that SOT can improve the PSNR of OT by about 2.6 dB, 2.7 dB and 1.3 dB, respectively, while achieving the best perception scores among the compared supervised and unsupervised methods. Particularly, on the three tasks, SOT significantly outperforms existing unsupervised methods and approaches the performance of state-of-the-art supervised methods.

Related papers

• Ultra: Unsupervised Cross-Task Optimization for Reliable Restoration Segmentation Collaboration under Adverse Weather [13.397348373903476]
  Unsupervised Domain Adaptation for Adverse Weather Semantic (UDA-ASS) aims to transfer semantic knowledge from labeled normal-weather images to unlabeled adverse environments.<n>We propose a novel Unsupervised Restoration-Segmentation Collaborative Learning Framework (Ultra), which reframes cross-task interaction as direction selection under uncertainty and causal effect estimation.<n>Our framework achieves better unsupervised restoration results than existing UDA-ASS restoration methods and generalizes to unsupervised restoration and object detection collaboration tasks.
  arXiv  Detail & Related papers  (2026-08-17T13:49:10Z)
• SCOUT: Self-Checking and Recovery-Aware Tool-Thought Agents for Ultra-Long Egocentric Video Reasoning [45.039898006163405]
  We develop SCOUT (Self-Checking Chain-Of-Tool-thought), a recovery-aware agentic framework for ultra-long egocentric video understanding.<n>SCOUT trades off exploitation (zoom-in) and exploration (region switching) to enable robust multi-hop reasoning over extremely long horizons.<n>We show that SCOUT achieves state-of-the-art results on ultra-long egocentric benchmarks, while remaining competitive on shorter-horizon long-video settings.
  arXiv  Detail & Related papers  (2026-08-08T06:48:08Z)
• Noise-Free One-Step LoRA for Task-Driven Image Restoration with Diffusion Priors [54.200377321709546]
  Degraded images not only reduce visual quality but also impair downstream high-level vision tasks.<n> Task-driven image restoration (TDIR) addresses this issue by jointly optimizing restoration quality and task performance.<n>We show that a deterministic, noise-free one-step forward pass with pretrained diffusion priors can substantially improve TDIR.<n>We also introduce a task-preserving GAN training strategy that improves perceptual quality without sacrificing task performance.
  arXiv  Detail & Related papers  (2026-07-28T07:51:44Z)
• Agentic Spatio-Temporal Grounding via Collaborative Reasoning [80.83158605034465]
  Temporal Video Grounding aims to retrieve thetemporal tube of a target object or person in a video given a text query.<n>We propose the Agentic Spatio-Temporal Grounder (ASTG) framework for the task of STVG towards an open-world and training-free scenario.<n>Specifically, two specialized agents SRA (Spatial Reasoning Agent) and TRA (Temporal Reasoning Agent) constructed leveraging on modern Multimoal Large Language Models (MLLMs)<n>Experiments on popular benchmarks demonstrate the superiority of the proposed approach where it outperforms existing weakly-supervised and zero-shot approaches by a margin
  arXiv  Detail & Related papers  (2026-02-10T10:16:27Z)
• Labels Matter More Than Models: Quantifying the Benefit of Supervised Time Series Anomaly Detection [56.302586730134806]
  Time series anomaly detection (TSAD) is a critical data mining task often constrained by label scarcity.<n>Current research predominantly focuses on Unsupervised Time-series Anomaly Detection.<n>This paper challenges the premise that architectural complexity is the optimal path for TSAD.
  arXiv  Detail & Related papers  (2025-11-20T08:32:49Z)
• Overthinking Reduction with Decoupled Rewards and Curriculum Data Scheduling [41.834250664485666]
  Large reasoning models generate excessively long reasoning paths without any performance benefit.<n>Existing solutions that penalize length often fail, inducing performance degradation.<n>We introduce a novel framework, DECS, built on our theoretical discovery of two previously unaddressed flaws in current length rewards.
  arXiv  Detail & Related papers  (2025-09-30T06:04:43Z)
• Supervised Optimism Correction: Be Confident When LLMs Are Sure [91.7459076316849]
  We establish a novel theoretical connection between supervised fine-tuning and offline reinforcement learning.<n>We show that the widely used beam search method suffers from unacceptable over-optimism.<n>We propose Supervised Optimism Correction, which introduces a simple yet effective auxiliary loss for token-level $Q$-value estimations.
  arXiv  Detail & Related papers  (2025-04-10T07:50:03Z)
• InvFussion: Bridging Supervised and Zero-shot Diffusion for Inverse Problems [76.39776789410088]
  This work introduces a framework that combines the strong performance of supervised approaches and the flexibility of zero-shot methods.<n>A novel architectural design seamlessly integrates the degradation operator directly into the denoiser.<n> Experimental results on the FFHQ and ImageNet datasets demonstrate state-of-the-art posterior-sampling performance.
  arXiv  Detail & Related papers  (2025-04-02T12:40:57Z)
• Offline Reinforcement Learning via Inverse Optimization [3.0586855806896054]
  We propose a novel offline Reinforcement Learning (ORL) algorithm for continuous state and action spaces. To mitigate the distribution shift commonly observed in ORL problems, we employ a robust and non-causal Model Predictive Control expert. Unlike the existing literature, our robust MPC expert enjoys an exact and tractable convex reformulation.
  arXiv  Detail & Related papers  (2025-02-27T12:11:44Z)
• Efficient Diffusion as Low Light Enhancer [63.789138528062225]
  Reflectance-Aware Trajectory Refinement (RATR) is a simple yet effective module to refine the teacher trajectory using the reflectance component of images. textbfReflectance-aware textbfDiffusion with textbfDistilled textbfTrajectory (textbfReDDiT) is an efficient and flexible distillation framework tailored for Low-Light Image Enhancement (LLIE)
  arXiv  Detail & Related papers  (2024-10-16T08:07:18Z)
• Unsupervised Variational Translator for Bridging Image Restoration and High-Level Vision Tasks [24.076965636237098]
  We propose an unsupervised learning method called textVariational textbfTranslator (VaT), which does not require retraining existing restoration and high-level vision networks.<n>VaT achieves the above optimization objective without requiring labels.<n>Experiments in dehazing and low-light enhancement for detection and classification show the superiority of our method over other state-of-the-art unsupervised counterparts.
  arXiv  Detail & Related papers  (2024-08-15T13:35:59Z)
• ACTRESS: Active Retraining for Semi-supervised Visual Grounding [52.08834188447851]
  A previous study, RefTeacher, makes the first attempt to tackle this task by adopting the teacher-student framework to provide pseudo confidence supervision and attention-based supervision. This approach is incompatible with current state-of-the-art visual grounding models, which follow the Transformer-based pipeline. Our paper proposes the ACTive REtraining approach for Semi-Supervised Visual Grounding, abbreviated as ACTRESS.
  arXiv  Detail & Related papers  (2024-07-03T16:33:31Z)
• Variational Delayed Policy Optimization [25.668512485348952]
  In environments with delayed observation, state augmentation by including actions within the delay window is adopted to retrieve Markovian property to enable reinforcement learning (RL) State-of-the-art (SOTA) RL techniques with Temporal-Difference (TD) learning frameworks often suffer from learning inefficiency, due to the significant expansion of the augmented state space with the delay. This work introduces a novel framework called Variational Delayed Policy Optimization (VDPO), which reformulates delayed RL as a variational inference problem.
  arXiv  Detail & Related papers  (2024-05-23T06:57:04Z)
• Preserving Silent Features for Domain Generalization [6.568921669414849]
  Self-supervised contrastive learning pre-trained models do not exhibit better generalization performance than supervised models pre-trained on the same dataset in the DG setting. We propose a simple yet effective method termed STEP (Silent Feature Preservation) to improve the generalization performance of the self-supervised contrastive learning pre-trained model.
  arXiv  Detail & Related papers  (2024-01-06T09:11:41Z)
• Adaptive trajectory-constrained exploration strategy for deep reinforcement learning [6.589742080994319]
  Deep reinforcement learning (DRL) faces significant challenges in addressing the hard-exploration problems in tasks with sparse or deceptive rewards and large state spaces. We propose an efficient adaptive trajectory-constrained exploration strategy for DRL. We conduct experiments on two large 2D grid world mazes and several MuJoCo tasks.
  arXiv  Detail & Related papers  (2023-12-27T07:57:15Z)
• Gradient constrained sharpness-aware prompt learning for vision-language models [99.74832984957025]
  This paper targets a novel trade-off problem in generalizable prompt learning for vision-language models (VLM) By analyzing the loss landscapes of the state-of-the-art method and vanilla Sharpness-aware Minimization (SAM) based method, we conclude that the trade-off performance correlates to both loss value and loss sharpness. We propose a novel SAM-based method for prompt learning, denoted as Gradient Constrained Sharpness-aware Context Optimization (GCSCoOp)
  arXiv  Detail & Related papers  (2023-09-14T17:13:54Z)
• Majorization-Minimization for sparse SVMs [46.99165837639182]
  Support Vector Machines (SVMs) were introduced for performing binary classification tasks, under a supervised framework, several decades ago. They often outperform other supervised methods and remain one of the most popular approaches in the machine learning arena. In this work, we investigate the training of SVMs through a smooth sparse-promoting-regularized squared hinge loss minimization.
  arXiv  Detail & Related papers  (2023-08-31T17:03:16Z)
• Efficient Deep Reinforcement Learning Requires Regulating Overfitting [91.88004732618381]
  We show that high temporal-difference (TD) error on the validation set of transitions is the main culprit that severely affects the performance of deep RL algorithms. We show that a simple online model selection method that targets the validation TD error is effective across state-based DMC and Gym tasks.
  arXiv  Detail & Related papers  (2023-04-20T17:11:05Z)
• Unsupervised Reward Shaping for a Robotic Sequential Picking Task from Visual Observations in a Logistics Scenario [1.4502611532302039]
  We focus on an unloading problem, typical of the logistics sector, modeled as a sequential pick-and-place task. Modern machine learning techniques have shown to work better than classic systems. We propose and theoretically motivate a novel Unsupervised Reward Shaping algorithm.
  arXiv  Detail & Related papers  (2022-09-25T23:30:14Z)
• Relation Extraction with Weighted Contrastive Pre-training on Distant Supervision [22.904752492573504]
  We propose a weighted contrastive learning method by leveraging the supervised data to estimate the reliability of pre-training instances. Experimental results on three supervised datasets demonstrate the advantages of our proposed weighted contrastive learning approach.
  arXiv  Detail & Related papers  (2022-05-18T07:45:59Z)
• Digging into Uncertainty in Self-supervised Multi-view Stereo [57.04768354383339]
  We propose a novel Uncertainty reduction Multi-view Stereo (UMVS) framework for self-supervised learning. Our framework achieves the best performance among unsupervised MVS methods, with competitive performance with its supervised opponents.
  arXiv  Detail & Related papers  (2021-08-30T02:53:08Z)
• WSSOD: A New Pipeline for Weakly- and Semi-Supervised Object Detection [75.80075054706079]
  We propose a weakly- and semi-supervised object detection framework (WSSOD) An agent detector is first trained on a joint dataset and then used to predict pseudo bounding boxes on weakly-annotated images. The proposed framework demonstrates remarkable performance on PASCAL-VOC and MSCOCO benchmark, achieving a high performance comparable to those obtained in fully-supervised settings.
  arXiv  Detail & Related papers  (2021-05-21T11:58:50Z)
• Learning Sparse Rewarded Tasks from Sub-Optimal Demonstrations [78.94386823185724]
  Imitation learning learns effectively in sparse-rewarded tasks by leveraging the existing expert demonstrations. In practice, collecting a sufficient amount of expert demonstrations can be prohibitively expensive. We propose Self-Adaptive Learning (SAIL) that can achieve (near) optimal performance given only a limited number of sub-optimal demonstrations.
  arXiv  Detail & Related papers  (2020-04-01T15:57:15Z)
• A Comprehensive Approach to Unsupervised Embedding Learning based on AND Algorithm [18.670975246545208]
  Unsupervised embedding learning aims to extract good representation from data without the need for any manual labels. This paper proposes a new unsupervised embedding approach, called Super-AND, which extends the current state-of-the-art model. Super-AND outperforms all existing approaches and achieves an accuracy of 89.2% on the image classification task for CIFAR-10.
  arXiv  Detail & Related papers  (2020-02-26T13:22:04Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.