Wulin Xie

Wulin Xie, Rui Dai, Ruidong Ding et al.

Image Quality Assessment (IQA) predicts perceptual quality scores consistent with human judgments. Recent RL-based IQA methods built on MLLMs focus on generating visual quality descriptions and scores, ignoring two key reliability limitations: (i) although the model's prediction stability varies significantly across training samples, existing GRPO-based methods apply uniform advantage weighting, thereby amplifying noisy signals from unstable samples in gradient updates; (ii) most works emphasize text-grounded reasoning over images while overlooking the model's visual perception ability of image content. In this paper, we propose Q-Hawkeye, an RL-based reliable visual policy optimization framework that redesigns the learning signal through unified Uncertainty-Aware Dynamic Optimization and Perception-Aware Optimization. Q-Hawkeye estimates predictive uncertainty using the variance of predicted scores across multiple rollouts and leverages this uncertainty to reweight each sample's update strength, stabilizing policy optimization. To strengthen perceptual reliability, we construct paired inputs of degraded images and their original images and introduce an Implicit Perception Loss that constrains the model to ground its quality judgments in genuine visual evidence. Extensive experiments demonstrate that Q-Hawkeye outperforms state-of-the-art methods and generalizes better across multiple datasets. Our dataset and code are available at https://github.com/AMAP-ML/Q-Hawkeye.

Yi-Fan Zhang, Tao Yu, Haochen Tian et al. · pku

Despite notable advancements in Multimodal Large Language Models (MLLMs), most state-of-the-art models have not undergone thorough alignment with human preferences. This gap exists because current alignment research has primarily achieved progress in specific areas (e.g., hallucination reduction), while the broader question of whether aligning models with human preferences can systematically enhance MLLM capability remains largely unexplored. To this end, we introduce MM-RLHF, a dataset containing $\mathbf{120k}$ fine-grained, human-annotated preference comparison pairs. This dataset represents a substantial advancement over existing resources, offering superior size, diversity, annotation granularity, and quality. Leveraging this dataset, we propose several key innovations to improve both the quality of reward models and the efficiency of alignment algorithms. Notably, we introduce a Critique-Based Reward Model, which generates critiques of model outputs before assigning scores, offering enhanced interpretability and more informative feedback compared to traditional scalar reward mechanisms. Additionally, we propose Dynamic Reward Scaling, a method that adjusts the loss weight of each sample according to the reward signal, thereby optimizing the use of high-quality comparison pairs. Our approach is rigorously evaluated across $\mathbf{10}$ distinct dimensions and $\mathbf{27}$ benchmarks, with results demonstrating significant and consistent improvements in model performance. Specifically, fine-tuning LLaVA-ov-7B with MM-RLHF and our alignment algorithm leads to a $\mathbf{19.5}$% increase in conversational abilities and a $\mathbf{60}$% improvement in safety. We have open-sourced the preference dataset, reward model, training and evaluation code, as well as reward modeling and safety benchmarks. For more details, please visit our project page: https://mm-rlhf.github.io.

Lian Zhao, Jie Wen, Xiaohuan Lu et al.

In real-world scenarios, multi-view multi-label learning often encounters the challenge of incomplete training data due to limitations in data collection and unreliable annotation processes. The absence of multi-view features impairs the comprehensive understanding of samples, omitting crucial details essential for classification. To address this issue, we present a task-augmented cross-view imputation network (TACVI-Net) for the purpose of handling partial multi-view incomplete multi-label classification. Specifically, we employ a two-stage network to derive highly task-relevant features to recover the missing views. In the first stage, we leverage the information bottleneck theory to obtain a discriminative representation of each view by extracting task-relevant information through a view-specific encoder-classifier architecture. In the second stage, an autoencoder based multi-view reconstruction network is utilized to extract high-level semantic representation of the augmented features and recover the missing data, thereby aiding the final classification task. Extensive experiments on five datasets demonstrate that our TACVI-Net outperforms other state-of-the-art methods.

Yang Shi, Huanqian Wang, Wulin Xie et al. · pku

Multimodal Large Language Models (MLLMs) have achieved considerable accuracy in Optical Character Recognition (OCR) from static images. However, their efficacy in video OCR is significantly diminished due to factors such as motion blur, temporal variations, and visual effects inherent in video content. To provide clearer guidance for training practical MLLMs, we introduce the MME-VideoOCR benchmark, which encompasses a comprehensive range of video OCR application scenarios. MME-VideoOCR features 10 task categories comprising 25 individual tasks and spans 44 diverse scenarios. These tasks extend beyond text recognition to incorporate deeper comprehension and reasoning of textual content within videos. The benchmark consists of 1,464 videos with varying resolutions, aspect ratios, and durations, along with 2,000 meticulously curated, manually annotated question-answer pairs. We evaluate 18 state-of-the-art MLLMs on MME-VideoOCR, revealing that even the best-performing model (Gemini-2.5 Pro) achieves an accuracy of only 73.7%. Fine-grained analysis indicates that while existing MLLMs demonstrate strong performance on tasks where relevant texts are contained within a single or few frames, they exhibit limited capability in effectively handling tasks that demand holistic video comprehension. These limitations are especially evident in scenarios that require spatio-temporal reasoning, cross-frame information integration, or resistance to language prior bias. Our findings also highlight the importance of high-resolution visual input and sufficient temporal coverage for reliable OCR in dynamic video scenarios.

Wulin Xie, Yi-Fan Zhang, Chaoyou Fu et al. · pku

Existing MLLM benchmarks face significant challenges in evaluating Unified MLLMs (U-MLLMs) due to: 1) lack of standardized benchmarks for traditional tasks, leading to inconsistent comparisons; 2) absence of benchmarks for mixed-modality generation, which fails to assess multimodal reasoning capabilities. We present a comprehensive evaluation framework designed to systematically assess U-MLLMs. Our benchmark includes: Standardized Traditional Task Evaluation. We sample from 12 datasets, covering 10 tasks with 30 subtasks, ensuring consistent and fair comparisons across studies." 2. Unified Task Assessment. We introduce five novel tasks testing multimodal reasoning, including image editing, commonsense QA with image generation, and geometric reasoning. 3. Comprehensive Model Benchmarking. We evaluate 12 leading U-MLLMs, such as Janus-Pro, EMU3, VILA-U, and Gemini2-flash, alongside specialized understanding (e.g., Claude-3.5-Sonnet) and generation models (e.g., DALL-E-3). Our findings reveal substantial performance gaps in existing U-MLLMs, highlighting the need for more robust models capable of handling mixed-modality tasks effectively. The code and evaluation data can be found in https://mme-unify.github.io/.

Wulin Xie, Lian Zhao, Jiang Long et al.

Multi-view multi-label classification (MvMLC) has recently garnered significant research attention due to its wide range of real-world applications. However, incompleteness in views and labels is a common challenge, often resulting from data collection oversights and uncertainties in manual annotation. Furthermore, the task of learning robust multi-view representations that are both view-consistent and view-specific from diverse views still a challenge problem in MvMLC. To address these issues, we propose a novel framework for incomplete multi-view multi-label classification (iMvMLC). Our method factorizes multi-view representations into two independent sets of factors: view-consistent and view-specific, and we correspondingly design a graph disentangling loss to fully reduce redundancy between these representations. Additionally, our framework innovatively decomposes consistent representation learning into three key sub-objectives: (i) how to extract view-shared information across different views, (ii) how to eliminate intra-view redundancy in consistent representations, and (iii) how to preserve task-relevant information. To this end, we design a robust task-relevant consistency learning module that collaboratively learns high-quality consistent representations, leveraging a masked cross-view prediction (MCP) strategy and information theory. Notably, all modules in our framework are developed to function effectively under conditions of incomplete views and labels, making our method adaptable to various multi-view and multi-label datasets. Extensive experiments on five datasets demonstrate that our method outperforms other leading approaches.

Xi Yang, Pai Peng, Wulin Xie et al.

Few-shot image classification remains a critical challenge in the field of computer vision, particularly in data-scarce environments. Existing methods typically rely on pre-trained visual-language models, such as CLIP. However, due to the modality gap, which is the inconsistent distribution of image and text features in the joint embedding space, directly using these features as class prototypes often leads to suboptimal performance. To address this issue, we propose a novel Cross-Modal Mapping (CMM) method. This method globally aligns image features with the text feature space through linear transformation and optimizes their local spatial relationships using triplet loss, thereby significantly enhancing cross-modal consistency. Experimental results show that compared to other methods, CMM simplifies the training process and demonstrates higher efficiency. Furthermore, CMM improves the average Top-1 accuracy by 1.06% on 11 benchmark datasets compared to methods that partially fine-tune the backbone, and it performs excellently on 4 distribution shift datasets. Notably, CMM effectively mitigates the modality gap in pre-trained models, enabling text features to serve as effective class prototypes for image features, thus providing an efficient and highly generalizable solution for few-shot learning.

Bingyan Nie, Wulin Xie, Jiang Long et al.

Recently, multi-view and multi-label classification have become significant domains for comprehensive data analysis and exploration. However, incompleteness both in views and labels is still a real-world scenario for multi-view multi-label classification. In this paper, we seek to focus on double missing multi-view multi-label classification tasks and propose our dual-level contrastive learning framework to solve this issue. Different from the existing works, which couple consistent information and view-specific information in the same feature space, we decouple the two heterogeneous properties into different spaces and employ contrastive learning theory to fully disentangle the two properties. Specifically, our method first introduces a two-channel decoupling module that contains a shared representation and a view-proprietary representation to effectively extract consistency and complementarity information across all views. Second, to efficiently filter out high-quality consistent information from multi-view representations, two consistency objectives based on contrastive learning are conducted on the high-level features and the semantic labels, respectively. Extensive experiments on several widely used benchmark datasets demonstrate that the proposed method has more stable and superior classification performance.