Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks75 from public data
- Proceedings of the 29th ACM International Conference on Multimedia922
A robust deep learning framework for real-world SR that randomly erases potential adversarial noises in the frequency domain of input images or features and is more insensitive to adversarial attacks and presents more stable SR results than existing models and defenses is proposed.
- Component Divide-and-Conquer for Real-World Image Super-Resolution419
This paper proposes a Component Divide-and-Conquer model and a Gradient-Weighted loss for SR, exploring the utility of guiding SR model with low-level image components, and proposes a feasible way to balance the difficulties of image components for SR.
- Min-Entropy Latent Model for Weakly Supervised Object Detection232
A min-entropy latent model (MELM) is proposed for weakly supervised object detection, unified with feature learning and optimized with a recurrent learning algorithm, which progressively transfers the weak supervision to object locations.
- Identity-Preserving Talking Face Generation with Landmark and Appearance Priors160
A novel Transformer-based landmark generator is devised to infer lip and jaw landmarks from the audio and a video rendering model is built to translate the generated landmarks into face images, which helps generate realistic and identity-preserving visual content.
- 123
- 3D Human Pose Machines with Self-supervised Learning97
A simple yet effective self-supervised correction mechanism to learn all intrinsic structures of human poses from abundant images and develops a 3D human pose machine, which jointly integrates the 2D spatial relationship, temporal smoothness of predictions and 3D geometric knowledge.
- NTIRE 2024 Challenge on Image Super-Resolution (×4): Methods and Results83
This paper reviews the NTIRE 2024 challenge on image super-resolution (×4), highlighting the solutions proposed and the outcomes obtained, and offers a comprehensive overview of current trends in this field.
- Trash to Treasure: Harvesting OOD Data with Cross-Modal Matching for Open-Set Semi-Supervised Learning81
This work proposes a novel cross-modal matching strategy to detect OOD samples that substantially lifts the performance on open-set SSL and outperforms the state-of-the-art by a large margin.
- Out-of-Candidate Rectification for Weakly Supervised Semantic Segmentation73
A group ranking-based Out-of-Candidate Rectification (OCR) mechanism in a plug-and-play fashion that can achieve remarkable performance gains on both Pascal VOC and MS COCO datasets with negligible extra training overhead, which jus-tifies the effectiveness and generality of OCR.
- 68
- 59
- Image Restoration Through Generalized Ornstein-Uhlenbeck Bridge58
The Generalized Ornstein-Uhlenbeck Bridge (GOUB) model is introduced, leveraging the natural mean-reverting property of the generalized OU process and further eliminating the variance of its steady-state distribution through the Doob's h-transform enabling the recovery of high-quality images from low-quality ones.
- IDF-CR: Iterative Diffusion Process for Divide-and-Conquer Cloud Removal in Remote-Sensing Images57
This work presents an iterative diffusion process for cloud removal (IDF-CR), which exhibits strong generative capabilities to achieve component divide-and-conquer (CDC) cloud removal and refine the stable diffusion by implementing ControlNet.
- SAM-COD+: SAM-Guided Unified Framework for Weakly-Supervised Camouflaged Object Detection52
This paper proposes a unified COD framework, termed SAM-COD, which is capable of supporting arbitrary weakly-supervised labels and employs a prompt adapter to handle scribbles as prompts based on SAM, and introduces response filter and semantic matcher modules to improve the quality of the masks obtained by SAM under COD prompts.
- 50
- 49
- Pedestrian Detection with Deep Convolutional Neural Network44
This paper proposes to cascade simple Aggregated Channel Features (ACF) and rich Deep Convolutional Neural Network (DCNN) features for efficient and effective pedestrian detection in complex scenes.
- 43
- AIM 2020 Challenge on Real Image Super-Resolution: Methods and Results43
This paper introduces the real image Super-Resolution challenge that was part of the Advances in Image Manipulation (AIM) workshop, held in conjunction with ECCV 2020, and gauges the state-of-the-art approaches for real image SR in terms of PSNR and SSIM.
- NTIRE 2022 Challenge on Super-Resolution and Quality Enhancement of Compressed Video: Dataset, Methods and Results41
This paper reviews the NTIRE 2022 Challenge on Super-Resolution and Quality Enhancement of Compressed Video and proposed the LDV 2.0 dataset, which includes theLDV dataset (240 videos) and 95 additional videos.
- 36
- Towards Real-World Burst Image Super-Resolution: Benchmark and Method34
A large-scale real-world burst super-resolution dataset, i.e., RealBSR, is established to explore the faithful reconstruction of image details from multiple frames and a Federated Burst Affinity network (FBAnet) is introduced to investigate non-trivial pixel-wise displacements among images under real- world image degradation.
- Masked Images Are Counterfactual Samples for Robust Fine-Tuning27
A novel fine-tuning method is proposed, which uses masked images as counterfactual samples that help improve the robustness of the fine- Tuning model, surpassing previous methods on the OOD performance.
- 27
- Dual Adversarial Adaptation for Cross-Device Real-World Image Super-Resolution25
This paper proposes an unsupervised domain adaptation mechanism for real-world SR, named Dual ADversarial Adaptation (DADA), which only requires LR images in the target domain with available real paired data from a source camera, and improves the model transferability from the source domain to the target.
- Real-World Image Super-Resolution by Exclusionary Dual-Learning24
This paper proposes a method, Real-World image Super-Resolution by Exclusionary Dual-Learning (RWSR-EDL) to address the feature diversity in perceptual- and L1- based cooperative learning and develops a noise-guidance data collection strategy to addressed the training time consumption in multiple datasets optimization.
- 23
- 20
- Multi-Task Classification and Segmentation for Explicable Capsule Endoscopy Diagnostics20
A multi-task framework, called the MTCSN, is proposed, to achieve joint learning of clearness degree (CD) and tissue semantic segmentation (TSS) for the first time and presents a new benchmark, named the Capsule-Endoscopy Crohn’s Disease dataset, which introduces the challenges faced in the real world.
- Diff-Mosaic: Augmenting Realistic Representations in Infrared Small Target Detection via Diffusion Prior19
The proposed Diff-Mosaic data augmentation method effectively alleviates the challenge of diversity and realism of data augmentation methods via diffusion prior, and introduces an enhancement network called Pixel-Prior, which generates highly coordinated and realistic Mosaic images by harmonizing pixels.
- 19
- Deductive Learning for Weakly-Supervised 3D Human Pose Estimation via Uncalibrated Cameras18
This paper proposes a Deductive Weakly-Supervised Learning (DWSL) for 3D human pose machine that introduces deductive reasoning to make an inference for the human pose from a view to another and develops a reconstruction loss to demonstrate what the model learns and infers is reliable.
- Scene Graph to Image Synthesis via Knowledge Consensus13
This paper delves into the causality of the adversarial generation process, and reason out a new principle to realize a simultaneous semantic disentanglement with an alignment on target and model distributions, named knowledge consensus, which determines a new graph-to-image generation framework.
- Progressively diffused networks for semantic visual parsing13
This work proposes Progressively Diffused Networks (PDNs) to deal with complex semantic parsing tasks and demonstrates the effectiveness of PDNs to substantially improve the performances of existing LSTM-based models.
- 13
- Open Set Domain Adaptation By Novel Class Discovery12
Self-supervised Class-Discovering Adapter (SCDA) is proposed that attempts to achieve Open Set Domain Adaptation by gradually discovering those implicit classes, then incorporating them to restructure the classifier and update the domain-adaptive features iteratively.
- 11
- DQnet: Cross-Model Detail Querying for Camouflaged Object Detection10
It is argued that partial activation is caused by the intrinsic characteristics of CNN, where the convolution operations produce local receptive fields and experience difficulty to capture long-range feature dependency among image regions.
- Deep CockTail Networks9
A theoretical analysis towards Deep CockTail Network (DCTN), a universal and flexibly-deployed framework to address the problems of transferable deep representations for visual domain adaptation, reveals that DCTN significantly boosts classification accuracies in MSDA and performs extraordinarily to resist negative transfers across different MSDA scenarios.
- 8
- 7
- 6
- Where Detectors Fail: Probing Generative Space for Generalizable AI-Generated Image Detection5
Experimental results across multiple benchmarks indicate that PROBE enhances generalization to unseen generators, resulting in more generalizable AIGI detection performance.
- Cross-Prompt Adversarial Attack on Segment Anything Model5
The problem of cross-prompt adversarial attacks on SAM is investigated, which considers whether the adversarial samples obtained by attacking with one or more prompts can lead to incorrect masks under other unseen test prompts and proposes Omni-Attack-SAM, an innovative and effective attack method to produce adversarial samples that are transferable to unseen prompts.
- Perceptual Image Enhancement by Relativistic Discriminant Learning With Cross-Scale Aggregated Representation5
An automatic method for image quality enhancement that relies on a multi-level connected generator and a pair-wise relativistic discriminator and can automatically enhance the low-quality image into a digital single lens reflex (DSLR) quality.
- Tumor Tissue Segmentation for Histopathological Images5
This study firstly summarizes a specific pipeline of tumor segmentation, then, hard example mining in tumor segmentsation is explored, and experiments are conducted for evaluating segmentation performance of the method.
- Unveiling Perceptual Artifacts: A Fine-Grained Benchmark for Interpretable AI-Generated Image Detection4
This work introduces a fine-grained benchmark towards eXplainable AI-Generated image Detection, named X-AIGD, which provides pixel-level, categorized annotations of perceptual artifacts, spanning low-level distortions, high-level semantics, and cognitive-level counterfactuals, to facilitate fine-grained interpretability evaluation and deeper insight into model decision-making processes.
- Decoder-Only LLMs are Better Controllers for Diffusion Models4
The experimental results show that the enhanced text-to-image diffusion models with the adapter module are superior to the stat-of-the-art models in terms of text-to-image generation quality and reliability.
- 4
- 4
- A Belief Based Correlated Topic Model for Semantic Region Analysis in Far-Field Video Surveillance Systems4
Experiments show that the proposed belief based correlated topic model (BCTM) outperforms the state-of-the-art methods on qualitative results of learning semantic regions.
- Towards Understanding the Robustness of Diffusion-Based Purification: A Stochastic Perspective3
This work assesses DBP robustness under a novel attack setting, Deterministic White-box, and pinpoint stochasticity as the main factor in DBP robustness, and shows that ADDT improves the robustness of DBP models.
- 3
- 3
- 3
- 3
- 2
- Can We Achieve Efficient Diffusion without Self-Attention? Distilling Self-Attention Into Convolutions2
ConvFusion is proposed to replace conventional self-attention modules with Pyramid Convolution Blocks ($Delta$ ConvBlocks) and achieves performance comparable to transformer-based counterparts while reducing computational cost by 6929× and surpassing LinFusion by 5.42× in efficiency-all without compromising generative fidelity.
- 2
- 1
- 1
- Language Generation as Optimal Control: Closed-Loop Diffusion in Latent Control Space–
This work reformulates language generation as a stochastic optimal control problem, providing a unified theoretical perspective to analyze autoregressive and diffusion models and explain their limitations in terms of combination of trajectory singularity, adjoint state vanishing, and gradient absence.
- Ground4D: Consistency-Aware 4D Reconstruction from Monocular Video–
By integrating foundation-level geometric priors into dynamic Gaussian optimization, Ground4D achieves stronger reconstruction fidelity and rendering performance, underscoring the role of geometry-grounded constraints in robust 4D scene modeling.
- –
- –
- –
- –
- –
- When Preference Labels Fall Short: Aligning Diffusion Models from Real Data–
This work adopts a data-centric perspective and study a curation strategy that treats real images as reference points and constructs preference signals by contrasting them with generated or perturbed samples, without requiring manually annotated preference pairs.
- –
- –
- FIRE: Fine Implicit Reconstruction Enhancement with Detailed Body Part Labels and Geometric Features–
- –
- –
- Generalizing Energy-based Generative ConvNets from Particle Evolution Perspective–
This paper establishes a connection between FRAME model and dynamic physics process and provides a generalized formulation of FRAME in discrete flow with a certain metric measure from particle perspective, and presents a Generalized GCN (GGCN), which introduces a hidden space mapping strategy and employs a normal distribution as hidden space for the reference distribution.
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.