Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks41 from public data
- AGIQA-3K: An Open Database for AI-Generated Image Quality Assessment263
This work extensively considers various popular AGI models, generated AGI through different prompts and model parameters, and collected subjective scores at the perceptual quality and text-to-image alignment, thus building the most comprehensive AGI subjective quality database AGIQA-3K so far.
- Q-Instruct: Improving Low-Level Visual Abilities for Multi-Modality Foundation Models218
The first dataset consisting of human natural language feedback on low-level vision, and a GPT-participated transformation to convert these feedbacks into a rich set of 200K instruction-response pairs, termed Q-Instruct are collected.
- GMS-3DQA: Projection-Based Grid Mini-patch Sampling for 3D Model Quality Assessment78
The experimental results show that the proposed GMS-3DQA outperforms existing state-of-the-art NR-3DQA methods on the point cloud quality assessment databases for both accuracy and efficiency.
- AIGIQA-20K: A Large Database for AI-Generated Image Quality Assessment65
The largest fine-grained AIGI subjective quality database to date with 20,000 AIGIs and 420,000 subjective ratings, known as AIGIQA-20K is released and it is anticipated that this large-scale quality database will inspire robust quality indicators for AIGIs and propel the evolution of AIGC for vision.
- Subjective-Aligned Dataset and Metric for Text-to-Video Quality Assessment60
This work proposes a novel transformer-based model for subjective-aligned Text-to-Video Quality Assessment (T2VQA), which extracts features from text-video alignment and video fidelity perspectives, then it leverages the ability of a large language model to give the prediction score.
- Large multimodal models evaluation: a survey57
By unifying general-specialized understanding and modality-specific generation evaluations, this survey clarifies the current landscape and provides guidance for future research in the LMM evaluation field.
- Learning Hazing to Dehazing: Towards Realistic Haze Generation for Real-World Image Dehazing56
A novel hazing-dehazing pipeline consisting of a Realistic Hazy Image Generation framework (HazeGen) and a Diffusion-based Dehazing framework (DiffDehaze), which harnesses robust generative diffusion priors of real-world hazy images embedded in a pre-trained text-to-image diffusion model.
- Quality Assessment in the Era of Large Models: A Survey55
A comprehensive review of the development of quality assessment in the era of large models focusing on two key aspects: the assessment of large models and the role of large models in assessment tasks.
- LMM-PCQA: Assisting Point Cloud Quality Assessment with LMM55
This study investigates the feasibility of imparting PCQA knowledge to LMMs through text supervision, and transforms quality labels into textual descriptions during the fine-tuning phase, enabling LMMs to derive quality rating logits from 2D projections of point clouds.
- Advancing Zero-Shot Digital Human Quality Assessment Through Text-Prompted Evaluation50
A zero-shot DHQA approach that focuses on no-reference scenarios to ensure generalization capabilities while mitigating database bias is proposed and the Digital Human Quality Index (DHQI) is introduced, which demonstrates significant improvements in zero-shot performance.
- 50
- MISC: Ultra-Low Bitrate Image Semantic Compression Driven by Large Multimodal Model47
Experimental results show that the proposed MISC is suitable for compressing both traditional Natural Sense Images (NSIs) and emerging AI-Generated Images (AIGIs) content and can achieve optimal consistency and perception results while saving 50% bitrate.
- Q-Eval-100K: Evaluating Visual Quality and Alignment Level for Text-to-Vision Content41
Experimental results indicate that the proposed Q-Eval-Score achieves superior performance on both visual quality and alignment, with strong generalization capabilities across other benchmarks, highlighting the significant value of the Q-EVAL-100K dataset.
- VQA 2 : Visual Question Answering for Video Quality Assessment33
This work provides a foundation and feasible approach for integrating low-level video quality assessment and understanding with LMMs, and presents the VQA² series models, which exceeds the renowned GPT-4o in visual quality understanding tasks while maintaining strong competitiveness in quality scoring tasks.
- 31
- Image Quality Assessment: From Human to Machine Preference28
The subjective preferences of machines are defined, including downstream tasks, test models, and evaluation metrics, and the Machine Preference Database is established, which contains 2.25M fine-grained annotations and 30k reference/distorted image pair instances.
- 28
- Benchmarking Multi-dimensional AIGC Video Quality Assessment: A Dataset and Unified Model27
The Unify Generated Video Quality Assessment (UGVQ) model is proposed, designed to accurately evaluate the multi-dimensional quality of AIGC videos, and established a benchmark for evaluating existing quality assessment metrics on the LGVQ dataset.
- HazeCLIP: Towards Language Guided Real-World Image Dehazing18
HazeCLIP is introduced, a language-guided adaptation framework designed to enhance the real-world performance of pre-trained dehazing networks, Inspired by the Contrastive Language-Image Pre-training (CLIP) model’s ability to distinguish between hazy and clean images, and leverage it to evaluate dehazing results.
- Towards All-in-One Medical Image Re-Identification12
This paper introduces a thorough benchmark and a unified model for medical image re-identification and proposes a novel Continuous Modality-based Parameter Adapter (ComPA), which condenses medical content into a continuous modality representation and dynamically adjusts the modality-agnostic model with modalityspecific parameters at runtime.
- T2I-Scorer: Quantitative Evaluation on Text-to-Image Generation via Fine-Tuned Large Multi-Modal Models12
The T2I-Scorer is introduced, a novel two-stage training methodology aimed at fine-tuning LMMs for T2I evaluation that has reached superior accuracy on T2I evaluation than all kinds of existing T2I metrics under zero-shot settings.
- G-Refine: A General Quality Refiner for Text-to-Image Generation11
G-Refine is introduced, a general image quality refiner designed to enhance low-quality images without compromising the integrity of high-quality ones, and significantly contributes to the practical application of contemporary T2I models, paving the way for their broader adoption.
- MM-PCQA+: Advancing Multi-Modal Learning for Point Cloud Quality Assessment10
This work presents an advanced no-reference Multi-Modal Point Cloud Quality Assessment (MM-PCQA+) metric, which surpasses all state-of-the-art methods and significantly advances beyond previous no-reference PCQA methods.
- Free-GVC: Towards Training-Free Extreme Generative Video Compression with Temporal Coherence8
This paper proposes Free-GVC, a training-free generative video compression framework that reformulates video coding as latent trajectory compression guided by a video diffusion prior, and introduces an Adaptive Quality Control module that dynamically constructs an online rate-perception surrogate model to predict the optimal diffusion step for each GOP.
- GeoX-Bench: Benchmarking Cross-View Geo-Localization and Pose Estimation Capabilities of Large Multimodal Models8
The benchmark demonstrate that while current LMMs achieve impressive performance in geo-localization tasks, their effectiveness declines significantly on the more complex pose estimation tasks, highlighting a critical area for future improvement, and instruction-tuning LMMs on the training data of GeoX-Bench can significantly improve the cross-view geo-sense abilities.
- Information Density Principle for MLLM Benchmarks8
A critical principle of Information Density is proposed, which examines how much insight a benchmark can provide for the development of MLLMs and shows that using the latest benchmarks in testing can provide more insight compared to previous ones, but there is still room for improvement in their information density.
- Optimizing Projection-Based Point Cloud Quality Assessment with Human Preferred Viewpoints Selection7
- SIQA: Toward Reliable Scientific Image Quality Assessment5
Experiments across representative multimodal large language models reveal a consistent discrepancy between scoring alignment and scientific understanding, suggesting that rating consistency alone may not reliably reflect scientific comprehension, underscoring the necessity of multidimensional evaluation for scientific image quality assessment.
- 3
- Embodied Image Quality Assessment for Robotic Intelligence2
This paper proposes the first Embodied Preference Database (EPD), which contains 12,500 distorted image annotations, and proposes a novel Multi-scale Attention Embodied Image Quality Assessment called MA-EIQA, which is the first no-reference IQA model designed for embodied robot.
- STAR : Bridging Statistical and Agentic Reasoning for Large Model Performance Prediction2
STARS is proposed, a framework that bridges data-driven STatistical expectations with knowledge-driven Agentic Reasoning and consistently outperforms all baselines on both score-based and rank-based metrics.
- AU-IQA: A Benchmark Dataset for Perceptual Quality Assessment of AI-Enhanced User-Generated Content2
AU-IQA, a benchmark dataset comprising 4,800 AI-UGC images produced by three representative enhancement types which include super-resolution, low-light enhancement, and denoising, is constructed and a comprehensive analysis of how well current approaches perform in assessing the perceptual quality of AI-UGC is provided.
- Towards a New Paradigm of Visual Signal Compression2
CMC- Bench, a benchmark of the cooperative performance of Image-to- Text (I2T) and Text-to-Image (T2I) models for image compression, proves that the combination of some I2T and T2I models has surpassed the most advanced visual signal codecs; meanwhile, it highlights where LMMs can be further optimized toward the compression task.
- 2
- 2
- Towards Characterizing Scientific Image Utility and Upgradability1
The SIU$^2$A-Benchmark is constructed, a dataset with expert annotations for error identification and repair, and the framework implements a two-stage evaluation protocol, which introduces two complementary dimensions for scientific image evaluation.
- Automated Safety Benchmarking: A Multi-agent Pipeline for LVLMs1
VLSafetyBencher is proposed, the first automated system for LVLM safety benchmarking, which introduces four collaborative agents: Data Preprocessing, Generation, Augmentation, and Selection agents to construct and select high-quality samples.
- Data Assessment for Embodied Intelligence1
This work introduces the first interpretable, data-driven algorithm to efficiently quantify dataset learnability without training, enabling researchers to assess a dataset's learnability immediately upon its release and proposes diversity entropy, a continuous measure that characterizes the amount of information contained in a dataset.
- –
- MirrorBench: Evaluating Self-centric Intelligence in MLLMs by Introducing a Mirror–
This study bridges psychological paradigms and embodied intelligence, offering a principled framework for evaluating the emergence of general intelligence in large models.
- –
Publication data from OpenAlex; 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.