Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks28 from public data
- Translating Natural Language to Planning Goals with Large-Language Models200
The empirical results on GPT 3.5 variants show that LLMs are much better suited towards translation rather than planning, and these models are promising for translation to structured planning languages, but care should be taken in their use.
- 95
- Sparse Sampling Transformer with Uncertainty-Driven Ranking for Unified Removal of Raindrops and Rain Streaks89
This paper proposes a Sparse Sampling Transformer based on Uncertainty-Driven Ranking, dubbed UDR-S2Former, which can adaptively sample relevant image degradation information to model underlying degradation relationships and clearly outperforms state-of-the-art methods for all benchmarks.
- Five A$^{+}$ Network: You Only Need 9K Parameters for Underwater Image Enhancement68
Thorough extensive experiments and comprehensive visual comparison, it is shown that FA$^{+}$Net outperforms previous approaches by obtaining state-of-the-art performance on multiple datasets while significantly reducing both parameter count and computational complexity.
- CVPR 2023 Text Guided Video Editing Competition66
This paper proposes a new dataset for text-guided video editing (TGVE), and runs a competition at CVPR to evaluate models on the TGVE dataset, and presents a retrospective on the competition and describes the winning method.
- 59
- 42
- HumanEdit: A High-Quality Human-Rewarded Dataset for Instruction-based Image Editing32
High-quality, human-rewarded dataset specifically designed for instruction-guided image editing, enabling precise and diverse image manipulations through open-form language instructions, set a new versatile benchmark for instructional image editing datasets.
- An Empirical Study of GPT-4o Image Generation Capabilities31
An empirical study of GPT-4o's image generation capabilities is conducted, benchmarking it against leading open-source and commercial models and identifying promising directions for future unified generative models, emphasizing the role of architectural design and data scaling.
- Rethinking the Design Space of Reinforcement Learning for Diffusion Models: On the Importance of Likelihood Estimation Beyond Loss Design20
This work shows that adopting an evidence lower bound (ELBO) based model likelihood estimator is the dominant factor enabling effective, efficient, and stable RL optimization, outweighing the impact of the specific policy-gradient loss functional.
- Integrating View Conditions for Image Synthesis20
A pioneering framework that integrates viewpoint information to enhance the control of image editing tasks, especially for interior design scenes is introduced, establishing a promising avenue for advancing image synthesis techniques and empowering precise object modifications while preserving the visual coherence of the entire composition.
- CPLFormer: Cross-scale Prototype Learning Transformer for Image Snow Removal17
This work proposes a novel approach, CPLFormer, which uses snow prototypes to own comprehensive clean scene understanding through learning from cross-scale features, outperforming convolutional network and vanilla transformer-based solutions.
- LaT: Latent Translation with Cycle-Consistency for Video-Text Retrieval16
This work presents a novel mechanism for learning the translation relationship from a source modality space $\mathcal{S}$ to a target modalitySpace without the need for a joint latent space, which bridges the gap between visual and textual domains.
- Sequential Affinity Learning for Video Restoration10
The proposed Sequential Affinity Learning Network (SALN), a novel approach for video restoration that eliminates inefficient temporal modeling operators and pixel-level feature alignment in the network architecture, is designed based on an affinity mechanism that establishes direct correspondences between the Query frame, degraded sequence, and restored frames in latent space.
- Prism: Efficient Test-Time Scaling via Hierarchical Search and Self-Verification for Discrete Diffusion Language Models9
This work proposes Prism (Pruning, Remasking, and Integrated Self-verification Method), an efficient TTS framework for dLLMs that achieves a favorable performance-efficiency trade-off, matching best-of-N performance with substantially fewer function evaluations (NFE).
- RSFDM-Net: Real-Time Spatial and Frequency Domains Modulation Network for Underwater Image Enhancement9
This work proposes a Real-time Spatial and Frequency Domains Modulation Network (RSFDM-Net) and introduces a Three-branch Feature Extraction (TFE) block in the primary net that processes images pixel by pixel to integrate the color information extended by the same channel.
- 8
- 8
- 7
- Evaluating Image Caption via Cycle-consistent Text-to-Image Generation6
CAMScore utilizes a text-to-image model to generate images from captions and subsequently evaluates these generated images against the original images, and designs a three-level evaluation framework for CAMScore that encompasses pixel-level, semantic-level, and objective-level perspectives.
- 4
- Masked Generative Transformer Is What You Need for Image Editing3
This work presents EditMGT, an MGT-based editing framework that is the first of its kind and achieves state-of-the-art image similarity on multiple benchmarks while delivering 6x faster editing, demonstrating that MGTs offer a compelling alternative to diffusion-based editing.
- dMLLM-TTS: Self-Verified and Efficient Test-Time Scaling for Diffusion Multi-Modal Large Language Models3
dMLLM-TTS is proposed, a novel framework operating on two complementary scaling axes: trajectory exploration scaling to enhance the diversity of generated hypotheses, and iterative refinement scaling for stable generation, which substantially improves generation quality while achieving up to 6x greater efficiency than linear search.
- Threshold-Guided Optimization for Visual Generative Models2
A threshold-guided alignment framework is proposed that replaces this oracle baseline with a data-driven global threshold estimated from empirical score statistics, enabling effective optimization directly from scalar feedback.
- PerceptionDLM: Parallel Region Perception with Multimodal Diffusion Language Models1
PerceptionDLM is the first to achieve parallel region caption and perception by leveraging the advantages of diffusion language models, and highlights the potential of multimodal diffusion language models for efficient, parallel visual perception.
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 2026-10-12. 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.