Multimodal Large Language Models struggle to maintain reliable performance under extreme real-world visual degradations, which impede their practical robustness. Existing robust MLLMs predominantly rely on implicit training/adaptation that focuses solely on visual encoder generalization, suffering from limited interpretability and isolated optimization. To overcome these limitations, we propose Robust-R1, a novel framework that explicitly models visual degradations through structured reasoning chains. Our approach integrates: (i) supervised fine-tuning for degradation-aware reasoning foundations, (ii) reward-driven alignment for accurately perceiving degradation parameters, and (iii) dynamic reasoning depth scaling adapted to degradation intensity. To facilitate this approach, we introduce a specialized 11K dataset featuring realistic degradations synthesized across four critical real-world visual processing stages, each annotated with structured chains connecting degradation parameters, perceptual influence, pristine semantic reasoning chain, and conclusion. Comprehensive evaluations demonstrate state-of-theart robustness: Robust-R1 outperforms all general and robust baselines on the real-world degradation benchmark R-Bench, while maintaining superior anti-degradation performance under multi-intensity adversarial degradations on MMMB, MMStar, and RealWorldQA.

Recent advances in multimodal LLMs and systems that use tools for long-video QA point to the promise of reasoning over hour-long episodes.However, many methods still compress content into lossy summaries or rely on limited toolsets, weakening temporal grounding and missing fine-grained cues.We propose a multi-agent framework in which a master LLM coordinates a grounding agent to localize question-relevant segments and a vision agent to extract targeted textual observations.The master agent plans with a step limit, and is trained with reinforcement learning to encourage concise, correct, and efficient multi-agent cooperation.This design helps the master agent focus on relevant clips via grounding, complements subtitles with visual detail, and yields interpretable trajectories.On our proposed LongTVQA and LongTVQA+ which are episode-level datasets aggregated from TVQA/TVQA+, our multiagent system significantly outperforms strong non-agent baselines.Experiments also show reinforcement learning further strengthens reasoning and planning for the trained agent.

Recent progress in generative diffusion models has greatly advanced text-to-video generation. While text-to-video models trained on large-scale, diverse datasets can produce varied outputs, these generations often deviate from user preferences, highlighting the need for preference alignment on pre-trained models. Although Direct Preference Optimization (DPO) [42] has demonstrated significant improvements in language and image generation [52], we pioneer its adaptation to video diffusion models and propose a VideoDPO pipeline by making several key adjustments. Unlike previous image alignment methods that focus solely on either (i) visual quality or (ii) semantic alignment between text and videos, we comprehensively consider both dimensions and construct a preference score accordingly, which we term the OmniScore. We design a pipeline to automatically collect preference pair data based on the proposed OmniScore and discover that re-weighting these pairs based on the score significantly impacts overall preference alignment. Our experiments demonstrate substantial improvements in both visual quality and semantic alignment, ensuring that no preference aspect is neglected. Code and data are available at https://videodpo.github.io/.

Multimodal large language models (MLLMs) have demonstrated extraordinary capabilities in conducting conversations based on image inputs. However, we observe that MLLMs exhibit a pronounced form of visual sycophantic behavior. While similar behavior has also been noted in text-based large language models (LLMs), it becomes significantly more prominent when MLLMs process image inputs. We refer to this phenomenon as the “sycophantic modality gap.” To better understand this issue, we further analyze the factors that contribute to the exacerbation of this gap. To mitigate the visual sycophantic behavior, we first experiment with naive supervised fine-tuning to help the MLLM resist misleading instructions from the user. However, we find that this approach also makes the MLLM overly resistant to corrective instructions (i.e., stubborn even if it is wrong). To alleviate this trade-off, we propose Sycophantic Reflective Tuning (SRT), which enables the MLLM to engage in reflective reasoning, allowing it to determine whether a user’s instruction is misleading or corrective before drawing a conclusion. After applying SRT, we observe a significant reduction in sycophantic behavior toward misleading instructions, without resulting in excessive stubbornness when receiving corrective instructions.

This paper presents ThinkDiff, a novel alignment paradigm that empowers text-to-image diffusion models with multimodal in-context understanding and reasoning capabilities by integrating the strengths of vision-language models (VLMs). Existing multimodal diffusion finetuning methods largely focus on pixel-level reconstruction rather than in-context reasoning, and are constrained by the complexity and limited availability of reasoning-based datasets. ThinkDiff addresses these challenges by leveraging vision-language training as a proxy task, aligning VLMs with the decoder of an encoder-decoder large language model (LLM) instead of a diffusion decoder. This proxy task builds on the observation that the LLM decoder shares the same input feature space with diffusion decoders that use the corresponding LLM encoder for prompt embedding. As a result, aligning VLMs with diffusion decoders can be simplified through alignment with the LLM decoder. Without complex training and datasets, ThinkDiff effectively unleashes understanding, reasoning, and composing capabilities in diffusion models. Experiments demonstrate that ThinkDiff significantly improves accuracy from 19.2% to 46.3% on the challenging CoBSAT benchmark for multimodal in-context reasoning generation, with only 5 hours of training on 4 A100 GPUs. Additionally, ThinkDiff demonstrates exceptional performance in composing multiple images and texts into logically coherent images. Project page: https://mizhenxing.github.io/ThinkDiff.

This paper proposes the first GAN inversion-based method for multi-class sketch-based image generation (MCSBIG). MC-SBIG is a challenging task that requires strong prior knowledge due to the significant domain gap between sketches and natural images. Existing learning-based approaches rely on a large-scale paired dataset to learn the mapping between these two image modalities. However, since the public paired sketch-photo data are scarce, it is struggling for learning-based methods to achieve satisfactory results. In this work, we introduce a new approach based on GAN inversion, which can utilize a powerful pretrained generator to facilitate image generation from a given sketch. Our GAN inversion-based method has two advantages: 1. it can freely take advantage of the prior knowledge of a pretrained image generator; 2. it allows the proposed model to focus on learning the mapping from a sketch to a low-dimension latent code, which is a much easier task than directly mapping to a high-dimension natural image. We also present a novel shape loss to improve generation quality further. Extensive experiments are conducted to show that our method can produce sketch-faithful and photo-realistic images and significantly outperform the baseline methods.

Humans can easily detect and segment moving objects simply by observing how they move, even without knowledge of object semantics. Inspired by this, we develop a zero-shot unsupervised approach for learning object segmentations. The model comprises two visual pathways: an appearance pathway that segments individual RGB images into coherent object regions, and a motion pathway that predicts the flow vector for each region between consecutive video frames. The two pathways jointly reconstruct a new representation called segment flow. This decoupled representation of appearance and motion is trained in a self-supervised manner to reconstruct one frame from another.When pretrained on an unlabeled video corpus, the model can be useful for a variety of applications, including 1) primary object segmentation from a single image in a zero-shot fashion; 2) moving object segmentation from a video with unsupervised test-time adaptation; 3) image semantic segmentation by supervised fine-tuning on a labeled image dataset. We demonstrate encouraging experimental results on all of these tasks using pretrained models.

Referring object detection and referring image segmentation are important tasks that require joint understanding of visual information and natural language. Yet there has been evidence that current benchmark datasets suffer from bias, and current state-of-the-art models cannot be easily evaluated on their intermediate reasoning process. To address these issues and complement similar efforts in visual question answering, we build CLEVR-Ref+, a synthetic diagnostic dataset for referring expression comprehension. The precise locations and attributes of the objects are readily available, and the referring expressions are automatically associated with functional programs. The synthetic nature allows control over dataset bias (through sampling strategy), and the modular programs enable intermediate reasoning ground truth without human annotators. In addition to evaluating several state-of-the-art models on CLEVR-Ref+, we also propose IEP-Ref, a module network approach that significantly outperforms other models on our dataset. In particular, we present two interesting and important findings using IEP-Ref: (1) the module trained to transform feature maps into segmentation masks can be attached to any intermediate module to reveal the entire reasoning process step-by-step; (2) even if all training data has at least one object referred, IEP-Ref can correctly predict no-foreground when presented with false-premise referring expressions. To the best of our knowledge, this is the first direct and quantitative proof that neural modules behave in the way they are intended. We will release data and code for CLEVR-Ref+.

Citation metadata extraction plays an important role in academic information retrieval and knowledge management. Current works on this task generally use rule-based, template-based or learning-based approaches but these methods usually either rely on handcrafted features or are limited with domains. Recently, neural networks have shown strong ability in addressing sequence labeling tasks.

This electronic document is a "live" template. The various components of your paper [title, text, heads, etc.] are Abstract-Object detection in natural scenes has been widely researched in the past decade, and many deep learning based methods have achieved good performance on this task. This paper focuses on how to transfer and refine those object detection approaches from natural scene images to documents images, and proposes a deep learning-based page object (e.g., tables, formulae, figures) detection method. On the basis of traditional Convolutional Neural Network (CNN) based object detection methods, we redesign the region proposal method, the training strategy, the network structure and replace the Non-Maximum Suppression (NMS) with a dynamic programming algorithm. The experimental results show that it is essential to adjust some modules of the natural scene object detection approaches in order to better process the document images. The proposed method also achieved better performance compared with existing page object detection methods.

With more and more scientific documents becoming available in PDF format, recognition of formulae in these PDF documents is of great significance. In this paper, we propose a symbol dominance based formulae recognition approach to recovering formulae structures by using the rich information extracted directly from PDF files. The hierarchical structure of formula is represented by relationship tree, and the tree is built recursively based on symbol dominance, which considers both the spatial layout of symbols and the typesetting conventions of mathematics. In addition, we propose a special character recognition method to identify the formula characters with multiple components or variable unicode. Repeatable and comparable experiments have been done over two large datasets, IM2LATEX-100K and PDFME-10K. Experimental results demonstrate that our method is more adaptive and practical for PDF documents compared with other two existing available formulae recognition systems, INFTY and WYGIWYS.