Reinforcement fine-tuning (RFT) is a proliferating paradigm for LMM training. Analogous to high-level reasoning tasks, RFT is similarly applicable to low-level vision domains, including image quality assessment (IQA). Existing RFT-based IQA methods typically use rule-based output rewards to verify the model's rollouts but provide no reward supervision for the "think” process, leaving its correctness and efficacy uncontrolled. Furthermore, these methods typically fine-tune directly on downstream IQA tasks without explicitly enhancing the model’s native low-level visual quality perception, which may constrain its performance upper bound. In response to these gaps, we propose the multi‐stage RFT IQA framework (Refine-IQA). In Stage-1, we build the Refine-Perception-20K dataset (with 12 main distortions, 20,907 locally-distorted images, and over 55K RFT samples) and design multi-task reward functions to strengthen the model’s visual quality perception. In Stage-2, targeting the quality scoring task, we introduce a probability difference reward involved strategy for "think" process supervision. The resulting Refine-IQA Series Models achieve outstanding performance on both perception and scoring tasks—and, notably, our paradigm activates a robust "think” (quality interpretating) capability that also attains exceptional results on the corresponding quality interpreting benchmark.

With the rising interest in research on Large Multi-modal Models (LMMs) for video understanding, many studies have emphasized general video comprehension capabilities, neglecting the systematic exploration into video quality understanding. To address this oversight, we introduce Q-Bench-Video in this paper, a new benchmark specifically designed to evaluate LMMs' proficiency in discerning video quality. a) To ensure video source diversity, Q-Bench-Video encompasses videos from natural scenes, AI-generated content (AIGC), and computer graphics (CG). b) Building on the traditional multiple-choice questions format with the Yes-or-No and What-How categories, we include Open-ended questions to better evaluate complex scenarios. Additionally, we incorporate the video pair quality comparison question to enhance comprehensiveness. c) Beyond the traditional Technical, Aesthetic, and Temporal distortions, we have expanded our evaluation aspects to include the dimension of AIGC distortions, which addresses the increasing demand for video generation. Finally, we collect a total of 2,378 question-answer pairs and test them on 12 open-source & 5 proprietary LMMs. Our findings indicate that while LMMs have a foundational understanding of perceptual video quality, their performance remains incomplete and imprecise, with a notable discrepancy compared to the performance of human beings. Through Q-Bench-Video, we seek to catalyze community interest, stimulate further research, and unlock the untapped potential of LMMs to close the gap in video quality understanding.

We introduce OBI-Bench, a holistic benchmark crafted to systematically evaluate large multi-modal models (LMMs) on whole-process oracle bone inscriptions (OBI) processing tasks demanding expert-level domain knowledge and deliberate cognition. OBI-Bench includes 5,523 meticulously collected diverse-sourced images, covering five key domain problems: recognition, rejoining, classification, retrieval, and deciphering. These images span centuries of archaeological findings and years of research by front-line scholars, comprising multi-stage font appearances from excavation to synthesis, such as original oracle bone, inked rubbings, oracle bone fragments, cropped single characters, and handprinted characters. Unlike existing benchmarks, OBI-Bench focuses on advanced visual perception and reasoning with OBI-specific knowledge, challenging LMMs to perform tasks akin to those faced by experts. The evaluation of 6 proprietary LMMs as well as 17 open-source LMMs highlights the substantial challenges and demands posed by OBI-Bench. Even the latest versions of GPT-4o, Gemini 1.5 Pro, and Qwen-VL-Max are still far from public-level humans in some fine-grained perception tasks. However, they perform at a level comparable to untrained humans in deciphering tasks, indicating remarkable capabilities in offering new interpretative perspectives and generating creative guesses. We hope OBI-Bench can facilitate the community to develop domain-specific multi-modal foundation models towards ancient language research and delve deeper to discover and enhance these untapped potentials of LMMs.

How to accurately and efficiently assess AI-generated images (AIGIs) remains a critical challenge for generative models. Given the high costs and extensive time commitments required for user studies, many researchers have turned towards employing large multi-modal models (LMMs) as AIGI evaluators, the precision and validity of which are still questionable. Furthermore, traditional benchmarks often utilize mostly natural-captured content rather than AIGIs to test the abilities of LMMs, leading to a noticeable gap for AIGIs. Therefore, we introduce **A-Bench** in this paper, a benchmark designed to diagnose *whether LMMs are masters at evaluating AIGIs*. Specifically, **A-Bench** is organized under two key principles: 1) Emphasizing both high-level semantic understanding and low-level visual quality perception to address the intricate demands of AIGIs. 2) Various generative models are utilized for AIGI creation, and various LMMs are employed for evaluation, which ensures a comprehensive validation scope. Ultimately, 2,864 AIGIs from 16 text-to-image models are sampled, each paired with question-answers annotated by human experts. We hope that **A-Bench** will significantly enhance the evaluation process and promote the generation quality for AIGIs.

The oracle bone inscription (OBI) recognition plays a significant role in understanding the history and culture of ancient China. However, the existing OBI datasets suffer from a long-tail distribution problem, leading to biased performance of OBI recognition models across majority and minority classes. With recent advancements in generative models, OBI synthesis-based data augmentation has become a promising avenue to expand the sample size of minority classes. Unfortunately, current OBI datasets lack large-scale structure-aligned image pairs for generative model training. To address these problems, we first present the Oracle-P15K, a structure-aligned OBI dataset for OBI generation and denoising, consisting of 14,542 images infused with domain knowledge from OBI experts. Second, we propose a diffusion model-based pseudo OBI generator, called OBIDiff, to achieve realistic and controllable OBI generation. Given a clean glyph image and a target rubbing-style image, it can effectively transfer the noise style of the original rubbing to the glyph image. Extensive experiments on OBI downstream tasks and user preference studies show the effectiveness of the proposed Oracle-P15K dataset and demonstrate that OBIDiff can accurately preserve inherent glyph structures while transferring authentic rubbing styles effectively. The dataset, code, and pre-trained models are available at https://github.com/LJHolyGround/Oracle-P15K.

Banding is a visually annoying artifact that frequently occurs along the chain of video acquisition, production, distribution, and display, showing a significant need for improvement in many fields. Thus far, efforts on banding removal are mainly knowledge-driven or merely learning on RGB space, which is either limited by domain knowledge or lacks the consideration for banding in chrominance channels. In this work, we propose a unified deep neural network that explicitly disentangles the luminance and chrominance channels, and simultaneously recovers intensity gradients and color discontinuity from detection-free measurement in an end-to-end manner. Our debanding model is comprised of a luminance restoration network (LR-Net) and a chrominance restoration network (CR-Net). Each of them follows an encoder-decoder architecture, where a cascade of residual blocks is employed to exploit hierarchical non-local features in spatial dimensions for more powerful feature representation. Moreover, we investigate the characteristics of banding artifacts and apply specific loss functions to guide the debanding in different channels, thus boosting the restoration performance. Both qualitative and quantitative experiments show that our model significantly surpasses the existing method in terms of all 7 metrics. Ultimately, our network trained on simulated data exhibits good adaptiveness under various compression scenarios, which further demonstrates the effectiveness of the proposed model.

The proliferation of Artificial Intelligence-Generated Images (AIGIs) has greatly expanded the Image Naturalness Assessment (INA) problem. Different from early definitions that mainly focus on tone-mapped images with limited distortions (e.g., exposure, contrast, and color reproduction), INA on AI-generated images is especially challenging as it owns more diverse contents and could be affected by factors from multiple perspectives, including low-level technical distortions and high-level rationality distortions. In this paper, we take the first step to benchmark and assess the visual naturalness of AI-generated images. First, we construct the AI-Generated Image Naturalness (AGIN) dataset by conducting a large-scale subjective study to collect human opinions on the overall naturalness as well as perceptions from the technical quality and rationality perspectives. AGIN verifies several insights for the first time that naturalness is universally and disparately affected by both technical and rational distortions, while its manifestations vary with different generation tasks. Second, to automatically assess the naturalness of AIGIs that align with human opinions, we propose the Joint Objective Image Naturalness evaluaTor (JOINT). Specifically, JOINT imitates human reasoning in naturalness evaluation by jointly learning technical and rationality features with several specific designs to guide model behavior from respective perspectives. Experiments demonstrate that JOINT significantly outperforms existing methods for providing more subjectively consistent results on naturalness assessment. The dataset can be accessed athttps://github.com/zijianchen98/AGIN.

In recent years, AI-driven video generation has gained significant attention due to great advancements in visual and language generative techniques. Consequently, there is a growing need for accurate Video Quality Assessment (VQA) metrics to evaluate the perceptual quality of AI-generated content (AIGC) videos and optimize video generation models. However, assessing the quality of AIGC videos remains a significant challenge because these videos often exhibit highly complex distortions, such as unnatural actions and irrational objects. To address this challenge, we systematically investigate the AIGC-VQA problem in this article, considering both subjective and objective quality assessment perspectives. For the subjective perspective, we construct the L arge-scale G enerated V ideo Q uality Assessment (LGVQ) dataset, consisting of \(2,\!808\) AIGC videos generated by six video generation models using 468 carefully curated text prompts. Unlike previous subjective VQA experiments, we evaluate the perceptual quality of AIGC videos from three critical dimensions: spatial quality, temporal quality, and text-video alignment, which hold utmost importance for current video generation techniques. For the objective perspective, we establish a benchmark for evaluating existing quality assessment metrics on the LGVQ dataset. Our findings show that current metrics perform poorly on this dataset, highlighting a gap in effective evaluation tools. To bridge this gap, we propose the U nify G enerated V ideo Q uality Assessment (UGVQ) model, designed to accurately evaluate the multi-dimensional quality of AIGC videos. The UGVQ model integrates the visual and motion features of videos with the textual features of their corresponding prompts, forming a unified quality-aware feature representation tailored to AIGC videos. Experimental results demonstrate that UGVQ achieves state-of-the-art performance on the LGVQ dataset across all three quality dimensions, validating its effectiveness as an accurate quality metric for AIGC videos. We hope that our benchmark can promote the development of AIGC-VQA studies. Both the LGVQ dataset and the UGVQ model are publicly available on https://github.com/zczhang-sjtu/UGVQ.git .

Just noticeable distortion (JND), representing the threshold of distortion in an image that is minimally perceptible to the human visual system (HVS), is crucial for image compression algorithms to achieve a trade-off between transmission bit rate and image quality. However, traditional JND prediction methods only rely on pixel-level or sub-band level features, lacking the ability to capture the impact of image content on JND. To bridge this gap, we propose a Semantic-Guided JND (SG-JND) network to leverage semantic information for JND prediction. In particular, SG-JND consists of three essential modules: the image preprocessing module extracts semantic-level patches from images, the feature extraction module extracts multi-layer features by utilizing the cross-scale attention layers, and the JND prediction module regresses the extracted features into the final JND value. Experimental results show that SG-JND achieves the state-of-the-art performance on two publicly available JND datasets, which demonstrates the effectiveness of SG-JND and highlight the significance of incorporating semantic information in JND assessment.

Banding artifact, as known as staircase-like contour, is a common quality annoyance that happens in compression, transmission, etc. scenarios, which largely affects the user’s quality of experience (QoE). The banding distortion typically appears as relatively small pixel-wise variations in smooth backgrounds, which is difficult to analyze in the spatial domain but easily reflected in the frequency domain. In this paper, we thereby study the banding artifact from the frequency aspect and propose a no-reference banding detection model to capture and evaluate banding artifacts, called the Frequency-Sensitive BANding Detector (FS-BAND). The proposed detector is able to generate a pixel-wise banding map with a perception correlated quality score. Experimental results show that the proposed FS-BAND method outperforms state-of-the-art image quality assessment (IQA) approaches with higher accuracy in banding classification task.

Assessing action quality is both imperative and challenging due to its significant impact on the quality of AI-generated videos, further complicated by the inherently ambiguous nature of actions within AI-generated video (AIGV). Current action quality assessment (AQA) algorithms predominantly focus on actions from real specific scenarios and are pre-trained with normative action features, thus rendering them inapplicable in AIGVs. To address these problems, we construct GAIA, a Generic AI-generated Action dataset, by conducting a large-scale subjective evaluation from a novel causal reasoning-based perspective, resulting in 971,244 ratings among 9,180 video-action pairs. Based on GAIA, we evaluate a suite of popular text-to-video (T2V) models on their ability to generate visually rational actions, revealing their pros and cons on different categories of actions. We also extend GAIA as a testbed to benchmark the AQA capacity of existing automatic evaluation methods. Results show that traditional AQA methods, action-related metrics in recent T2V benchmarks, and mainstream video quality methods perform poorly with an average SRCC of 0.454, 0.191, and 0.519, respectively, indicating a sizable gap between current models and human action perception patterns in AIGVs. Our findings underscore the significance of action quality as a unique perspective for studying AIGVs and can catalyze progress towards methods with enhanced capacities for AQA in AIGVs.

Banding, also known as staircase-like contours, frequently occurs in flat areas of images/videos processed by compression or quantization algorithms. As undesirable artifacts, banding destroys the original image structure, thus inevitably degrading users’ quality of experience (QoE). In this paper, we systematically investigate the banding image quality assessment (IQA) problem, aiming to detect the image banding artifacts and evaluate their perceptual visual quality. Considering that the existing image banding databases only contain limited content sources and banding generation methods, and lack perceptual quality labels (i.e. mean opinion scores), we first build the largest banding IQA database so far, namedBanding Artifact Noticeable Database (BAND-2k), which consists of 2,000 banding images generated by 15 compression and quantization schemes. A total of 23 workers participated in the subjective IQA experiment, yielding over 214,000 patch-level banding class labels and 44,371 reliable image-level quality rating scores. Subsequently, we develop an effective no-reference (NR) banding evaluator for banding detection and quality assessment by leveraging frequency characteristics of banding artifacts. To be more specific, a dual convolutional neural network (CNN) is employed to concurrently learn the feature representation from the high-frequency and low-frequency maps, thereby enhancing the ability to discern banding artifacts. The quality score of a banding image is generated by pooling the banding detection maps masked by the spatial frequency filters. The experimental results demonstrate that our banding evaluator achieves remarkably high accuracy in banding detection and also exhibits high SRCC and PLCC results with the perceptual quality labels, even without directly learning a regression model for banding quality evaluation. These findings unveil the strong correlations between the intensity of banding artifacts and the perceptual visual quality, thus validating the necessity of banding quality assessment. The BAND-2k database and the proposed banding evaluator are available at https://github.com/zijianchen98/BAND-2k.

Nowadays, most three-dimensional model quality assessment (3DQA) methods have been aimed at improving accuracy. However, little attention has been paid to the computational cost and inference time required for practical applications. Model-based 3DQA methods extract features directly from the 3D models, which are characterized by their high degree of complexity. As a result, many researchers are inclined towards utilizing projection-based 3DQA methods. Nevertheless, previous projection-based 3DQA methods directly extract features from multi-projections to ensure quality prediction accuracy, which calls for more resource consumption and inevitably leads to inefficiency. Thus, in this article, we address this challenge by proposing a no-reference (NR) projection-based G rid M ini-patch S ampling 3D Model Q uality A ssessment (GMS-3DQA) method. The projection images are rendered from six perpendicular viewpoints of the 3D model to cover sufficient quality information. To reduce redundancy and inference resources, we propose a multi-projection grid mini-patch sampling strategy (MP-GMS), which samples grid mini-patches from the multi-projections and forms the sampled grid mini-patches into one quality mini-patch map (QMM). The Swin-Transformer tiny backbone is then used to extract quality-aware features from the QMMs. 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. The efficiency analysis reveals that the proposed GMS-3DQA requires far less computational resources and inference time than other 3DQA competitors. The code is available at https://github.com/zzc-1998/GMS-3DQA .