Analyzing Reasoning Consistency in Large Multimodal Models under Cross-Modal Conflicts

Large Multimodal Models (LMMs) have demonstrated impressive capabilities in video reasoning via Chain-of-Thought (CoT). However, the robustness of their reasoning chains remains questionable. In this paper, we identify a critical failure mode termed textual inertia, where once a textual hallucination occurs in the thinking process, models tend to blindly adhere to the erroneous text while neglecting conflicting visual evidence. To systematically investigate this, we propose the LogicGraph Perturbation Protocol that structurally injects perturbations into the reasoning chains of diverse LMMs spanning both native reasoning architectures and prompt-driven paradigms to evaluate their self-reflection capabilities. The results reveal that models successfully self-correct in less than 10% of cases and predominantly succumb to blind textual error propagation. To mitigate this, we introduce Active Visual-Context Refinement, a training-free inference paradigm which orchestrates an active visual re-grounding mechanism to enforce fine-grained verification coupled with an adaptive context refinement strategy to summarize and denoise the reasoning history. Experiments demonstrate that our approach significantly stifles hallucination propagation and enhances reasoning robustness.

Semantics and Content Matter: Towards Multi-Prior Hierarchical Mamba for Image Deraining

Rain significantly degrades the performance of computer vision systems, particularly in applications like autonomous driving and video surveillance. While existing deraining methods have made considerable progress, they often struggle with fidelity of semantic and spatial details. To address these limitations, we propose the Multi-Prior Hierarchical Mamba (MPHM) network for image deraining. This novel architecture synergistically integrates macro-semantic textual priors (CLIP) for task-level semantic guidance and micro-structural visual priors (DINOv2) for scene-aware structural information. To alleviate potential conflicts between heterogeneous priors, we devise a progressive Priors Fusion Injection (PFI) that strategically injects complementary cues at different decoder levels. Meanwhile, we equip the backbone network with an elaborate Hierarchical Mamba Module (HMM) to facilitate robust feature representation, featuring a Fourier-enhanced dual-path design that concurrently addresses global context modeling and local detail recovery. Comprehensive experiments demonstrate MPHM's state-of-the-art performance, achieving a 0.57 dB PSNR gain on the Rain200H dataset while delivering superior generalization on real-world rainy scenarios.

Leveraging Static Relationships for Intra-Type and Inter-Type Message Passing in Video Question Answering

Video Question Answering (VideoQA) is an important research direction in the field of artificial intelligence, enabling machines to understand video content and perform reasoning and answering based on natural language questions. Although methods based on static relationship reasoning have made certain progress, there are still deficiencies in the accuracy of static relationship recognition and representation, and they have not fully utilized the static relationship information in videos for in-depth reasoning and analysis. Therefore, this paper proposes a reasoning method for intra-type and inter-type message passing based on static relationships. This method constructs a dual graph for intra-type message passing reasoning and builds a heterogeneous graph based on static relationships for inter-type message passing reasoning. The intra-type message passing reasoning model captures the neighborhood information of targets and relationships related to the question in the dual graph, updating the dual graph to obtain intra-type clues for answering the question. The inter-type message passing reasoning model captures the neighborhood information of targets and relationships from different categories related to the question in the heterogeneous graph, updating the heterogeneous graph to obtain inter-type clues for answering the question. Finally, the answers are inferred by combining the intra-type and inter-type clues based on static relationships. Experimental results on the ANetQA and Next-QA datasets demonstrate the effectiveness of this method.

Unbiased Scene Graph Generation by Type-Aware Message Passing on Heterogeneous and Dual Graphs

Although great progress has been made in the research of unbiased scene graph generation, issues still hinder improving the predictive performance of both head and tail classes. An unbiased scene graph generation (TA-HDG) is proposed to address these issues. For modeling interactive and non-interactive relations, the Interactive Graph Construction is proposed to model the dependence of relations on objects by combining heterogeneous and dual graph, when modeling relations between multiple objects. It also implements a subject-object pair selection strategy to reduce meaningless edges. Moreover, the Type-Aware Message Passing enhances the understanding of complex interactions by capturing intra- and inter-type context in the Intra-Type and Inter-Type stages. The Intra-Type stage captures the semantic context of inter-relaitons and inter-objects. On this basis, the Inter-Type stage captures the context between objects and relations for interactive and non-interactive relations, respectively. Experiments on two datasets show that TA-HDG achieves improvements in the metrics of R@K and mR@K, which proves that TA-HDG can accurately predict the tail class while maintaining the competitive performance of the head class.

Neural-Symbolic VideoQA: Learning Compositional Spatio-Temporal Reasoning for Real-world Video Question Answering

Compositional spatio-temporal reasoning poses a significant challenge in the field of video question answering (VideoQA). Existing approaches struggle to establish effective symbolic reasoning structures, which are crucial for answering compositional spatio-temporal questions. To address this challenge, we propose a neural-symbolic framework called Neural-Symbolic VideoQA (NS-VideoQA), specifically designed for real-world VideoQA tasks. The uniqueness and superiority of NS-VideoQA are two-fold: 1) It proposes a Scene Parser Network (SPN) to transform static-dynamic video scenes into Symbolic Representation (SR), structuralizing persons, objects, relations, and action chronologies. 2) A Symbolic Reasoning Machine (SRM) is designed for top-down question decompositions and bottom-up compositional reasonings. Specifically, a polymorphic program executor is constructed for internally consistent reasoning from SR to the final answer. As a result, Our NS-VideoQA not only improves the compositional spatio-temporal reasoning in real-world VideoQA task, but also enables step-by-step error analysis by tracing the intermediate results. Experimental evaluations on the AGQA Decomp benchmark demonstrate the effectiveness of the proposed NS-VideoQA framework. Empirical studies further confirm that NS-VideoQA exhibits internal consistency in answering compositional questions and significantly improves the capability of spatio-temporal and logical inference for VideoQA tasks.