arXiv subjects

Zhuotao Tian

Publications and source records attributed to Zhuotao Tian.

At least 55 records · Page 3Linked to original sources

arXiv2026

FlashVID: Efficient Video Large Language Models via Training-free Tree-based Spatiotemporal Token Merging

Although Video Large Language Models (VLLMs) have shown remarkable capabilities in video understanding, they are required to process high volumes of visual tokens, causing significant computational inefficiency. Existing VLLMs acceleration frameworks usually compress spatial and temporal redundancy independently, which overlooks the spatiotemporal relationships, thereby leading to suboptimal spatiotemporal compression. The highly correlated visual features are likely to change in spatial position, scale, orientation, and other attributes over time due to the dynamic nature of video. Building on this insight, we introduce FlashVID, a training-free inference acceleration framework for VLLMs. Specifically, FlashVID utilizes Attention and Diversity-based Token Selection (ADTS) to select the most representative tokens for basic video representation, then applies Tree-based Spatiotemporal Token Merging (TSTM) for fine-grained spatiotemporal redundancy elimination. Extensive experiments conducted on three representative VLLMs across five video understanding benchmarks demonstrate the effectiveness and generalization of our method. Notably, by retaining only 10% of visual tokens, FlashVID preserves 99.1% of the performance of LLaVA-OneVision. Consequently, FlashVID can serve as a training-free and plug-and-play module for extending long video frames, which enables a 10x increase in video frame input to Qwen2.5-VL, resulting in a relative improvement of 8.6% within the same computational budget. Code is available at https://github.com/Fanziyang-v/FlashVID.

cs.CV↗

arXiv2025

Less Is More, but Where? Dynamic Token Compression via LLM-Guided Keyframe Prior

Recent advances in Video Large Language Models (VLLMs) have achieved remarkable video understanding capabilities, yet face critical efficiency bottlenecks due to quadratic computational growth with lengthy visual token sequences of long videos. While existing keyframe sampling methods can improve temporal modeling efficiency, additional computational cost is introduced before feature encoding, and the binary frame selection paradigm is found suboptimal. Therefore, in this work, we propose Dynamic Token compression via LLM-guided Keyframe prior (DyToK), a training-free paradigm that enables dynamic token compression by harnessing VLLMs' inherent attention mechanisms. Our analysis reveals that VLLM attention layers naturally encoding query-conditioned keyframe priors, by which DyToK dynamically adjusts per-frame token retention ratios, prioritizing semantically rich frames while suppressing redundancies. Extensive experiments demonstrate that DyToK achieves state-of-the-art efficiency-accuracy tradeoffs. DyToK shows plug-and-play compatibility with existing compression methods, such as VisionZip and FastV, attaining 4.3x faster inference while preserving accuracy across multiple VLLMs, such as LLaVA-OneVision and Qwen2.5-VL. Code is available at https://github.com/yu-lin-li/DyToK .

cs.CV↗

arXiv2025

Context-Aware Hierarchical Learning: A Two-Step Paradigm towards Safer LLMs

Large Language Models (LLMs) have emerged as powerful tools for diverse applications. However, their uniform token processing paradigm introduces critical vulnerabilities in instruction handling, particularly when exposed to adversarial scenarios. In this work, we identify and propose a novel class of vulnerabilities, termed Tool-Completion Attack (TCA), which exploits function-calling mechanisms to subvert model behavior. To evaluate LLM robustness against such threats, we introduce the Tool-Completion benchmark, a comprehensive security assessment framework, which reveals that even state-of-the-art models remain susceptible to TCA, with surprisingly high attack success rates. To address these vulnerabilities, we introduce Context-Aware Hierarchical Learning (CAHL), a sophisticated mechanism that dynamically balances semantic comprehension with role-specific instruction constraints. CAHL leverages the contextual correlations between different instruction segments to establish a robust, context-aware instruction hierarchy. Extensive experiments demonstrate that CAHL significantly enhances LLM robustness against both conventional attacks and the proposed TCA, exhibiting strong generalization capabilities in zero-shot evaluations while still preserving model performance on generic tasks. Our code is available at https://github.com/S2AILab/CAHL.

cs.CR↗

arXiv2025

SemanticVLA: Semantic-Aligned Sparsification and Enhancement for Efficient Robotic Manipulation

Vision-Language-Action (VLA) models have advanced in robotic manipulation, yet practical deployment remains hindered by two key limitations: 1) perceptual redundancy, where irrelevant visual inputs are processed inefficiently, and 2) superficial instruction-vision alignment, which hampers semantic grounding of actions. In this paper, we propose SemanticVLA, a novel VLA framework that performs Semantic-Aligned Sparsification and Enhancement for Efficient Robotic Manipulation. Specifically: 1) To sparsify redundant perception while preserving semantic alignment, Semantic-guided Dual Visual Pruner (SD-Pruner) performs: Instruction-driven Pruner (ID-Pruner) extracts global action cues and local semantic anchors in SigLIP; Spatial-aggregation Pruner (SA-Pruner) compacts geometry-rich features into task-adaptive tokens in DINOv2. 2) To exploit sparsified features and integrate semantics with spatial geometry, Semantic-complementary Hierarchical Fuser (SH-Fuser) fuses dense patches and sparse tokens across SigLIP and DINOv2 for coherent representation. 3) To enhance the transformation from perception to action, Semantic-conditioned Action Coupler (SA-Coupler) replaces the conventional observation-to-DoF approach, yielding more efficient and interpretable behavior modeling for manipulation tasks. Extensive experiments on simulation and real-world tasks show that SemanticVLA sets a new SOTA in both performance and efficiency. SemanticVLA surpasses OpenVLA on LIBERO benchmark by 21.1% in success rate, while reducing training cost and inference latency by 3.0-fold and 2.7-fold.SemanticVLA is open-sourced and publicly available at https://github.com/JiuTian-VL/SemanticVLA

cs.CV↗

arXiv2025

CalibCLIP: Contextual Calibration of Dominant Semantics for Text-Driven Image Retrieval

Existing Visual Language Models (VLMs) suffer structural limitations where a few low contribution tokens may excessively capture global semantics, dominating the information aggregation process and suppressing the discriminative features in text-driven image retrieval tasks. To address this, we introduce \textbf{CalibCLIP}, a training-free method designed to calibrate the suppressive effect of dominant tokens. Specifically, in the visual space, we propose the Contrastive Visual Enhancer (CVE), which decouples visual features into target and low information regions. Subsequently, it identifies dominant tokens and dynamically suppresses their representations.In the textual space, we introduce the Discriminative Concept Calibrator (DCC), which aims to differentiate between general and discriminative concepts within the text query. By mitigating the challenges posed by generic concepts and improving the representations of discriminative concepts, DCC strengthens the differentiation among similar samples. Finally, extensive experiments demonstrate consistent improvements across seven benchmarks spanning three image retrieval tasks, underscoring the effectiveness of CalibCLIP. Code is available at: https://github.com/kangbin98/CalibCLIP

cs.CV↗

arXiv2025

Memory Forcing: Spatio-Temporal Memory for Consistent Scene Generation on Minecraft

Autoregressive video diffusion models have proved effective for world modeling and interactive scene generation, with Minecraft gameplay as a representative application. To faithfully simulate play, a model must generate natural content while exploring new scenes and preserve spatial consistency when revisiting explored areas. Under limited computation budgets, it must compress and exploit historical cues within a finite context window, which exposes a trade-off: Temporal-only memory lacks long-term spatial consistency, whereas adding spatial memory strengthens consistency but may degrade new scene generation quality when the model over-relies on insufficient spatial context. We present Memory Forcing, a learning framework that pairs training protocols with a geometry-indexed spatial memory. Hybrid Training exposes distinct gameplay regimes, guiding the model to rely on temporal memory during exploration and incorporate spatial memory for revisits. Chained Forward Training extends autoregressive training with model rollouts, where chained predictions create larger pose variations and encourage reliance on spatial memory for maintaining consistency. Point-to-Frame Retrieval efficiently retrieves history by mapping currently visible points to their source frames, while Incremental 3D Reconstruction maintains and updates an explicit 3D cache. Extensive experiments demonstrate that Memory Forcing achieves superior long-term spatial consistency and generative quality across diverse environments, while maintaining computational efficiency for extended sequences.

cs.CV↗

arXiv2025

Generalized Decoupled Learning for Enhancing Open-Vocabulary Dense Perception

Dense visual perception tasks have been constrained by their reliance on predefined categories, limiting their applicability in real-world scenarios where visual concepts are unbounded. While Vision-Language Models (VLMs) like CLIP have shown promise in open-vocabulary tasks, their direct application to dense perception often leads to suboptimal performance due to limitations in local feature representation. In this work, we present our observation that CLIP's image tokens struggle to effectively aggregate information from spatially or semantically related regions, resulting in features that lack local discriminability and spatial consistency. To address this issue, we propose DeCLIP, a novel framework that enhances CLIP by decoupling the self-attention module to obtain ``content'' and ``context'' features respectively. \revise{The context features are enhanced by jointly distilling semantic correlations from Vision Foundation Models (VFMs) and object integrity cues from diffusion models, thereby enhancing spatial consistency. In parallel, the content features are aligned with image crop representations and constrained by region correlations from VFMs to improve local discriminability. Extensive experiments demonstrate that DeCLIP establishes a solid foundation for open-vocabulary dense perception, consistently achieving state-of-the-art performance across a broad spectrum of tasks, including 2D detection and segmentation, 3D instance segmentation, video instance segmentation, and 6D object pose estimation.} Code is available at https://github.com/xiaomoguhz/DeCLIP

cs.CV↗

arXiv2025

Perception, Reason, Think, and Plan: A Survey on Large Multimodal Reasoning Models

Reasoning lies at the heart of intelligence, shaping the ability to make decisions, draw conclusions, and generalize across domains. In artificial intelligence, as systems increasingly operate in open, uncertain, and multimodal environments, reasoning becomes essential for enabling robust and adaptive behavior. Large Multimodal Reasoning Models (LMRMs) have emerged as a promising paradigm, integrating modalities such as text, images, audio, and video to support complex reasoning capabilities and aiming to achieve comprehensive perception, precise understanding, and deep reasoning. As research advances, multimodal reasoning has rapidly evolved from modular, perception-driven pipelines to unified, language-centric frameworks that offer more coherent cross-modal understanding. While instruction tuning and reinforcement learning have improved model reasoning, significant challenges remain in omni-modal generalization, reasoning depth, and agentic behavior. To address these issues, we present a comprehensive and structured survey of multimodal reasoning research, organized around a four-stage developmental roadmap that reflects the field's shifting design philosophies and emerging capabilities. First, we review early efforts based on task-specific modules, where reasoning was implicitly embedded across stages of representation, alignment, and fusion. Next, we examine recent approaches that unify reasoning into multimodal LLMs, with advances such as Multimodal Chain-of-Thought (MCoT) and multimodal reinforcement learning enabling richer and more structured reasoning chains. Finally, drawing on empirical insights from challenging benchmarks and experimental cases of OpenAI O3 and O4-mini, we discuss the conceptual direction of native large multimodal reasoning models (N-LMRMs), which aim to support scalable, agentic, and adaptive reasoning and planning in complex, real-world environments.

cs.CV↗

arXiv2025

Edit360: 2D Image Edits to 3D Assets from Any Angle

Recent advances in diffusion models have significantly improved image generation and editing, but extending these capabilities to 3D assets remains challenging, especially for fine-grained edits that require multi-view consistency. Existing methods typically restrict editing to predetermined viewing angles, severely limiting their flexibility and practical applications. We introduce Edit360, a tuning-free framework that extends 2D modifications to multi-view consistent 3D editing. Built upon video diffusion models, Edit360 enables user-specific editing from arbitrary viewpoints while ensuring structural coherence across all views. The framework selects anchor views for 2D modifications and propagates edits across the entire 360-degree range. To achieve this, Edit360 introduces a novel Anchor-View Editing Propagation mechanism, which effectively aligns and merges multi-view information within the latent and attention spaces of diffusion models. The resulting edited multi-view sequences facilitate the reconstruction of high-quality 3D assets, enabling customizable 3D content creation.

cs.GR↗

arXiv2025

Enhancing Spatial Reasoning in Multimodal Large Language Models through Reasoning-based Segmentation

Recent advances in point cloud perception have demonstrated remarkable progress in scene understanding through vision-language alignment leveraging large language models (LLMs). However, existing methods may still encounter challenges in handling complex instructions that require accurate spatial reasoning, even if the 3D point cloud data provides detailed spatial cues such as size and position for identifying the targets. To tackle this issue, we propose Relevant Reasoning Segmentation (R$^2$S), a reasoning-based segmentation framework. The framework emulates human cognitive processes by decomposing spatial reasoning into two sequential stages: first identifying relevant elements, then processing instructions guided by their associated visual priors. Furthermore, acknowledging the inadequacy of existing datasets in complex reasoning tasks, we introduce 3D ReasonSeg, a reasoning-based segmentation dataset comprising 25,185 training samples and 3,966 validation samples with precise annotations. Both quantitative and qualitative experiments demonstrate that the R$^2$S and 3D ReasonSeg effectively endow 3D point cloud perception with stronger spatial reasoning capabilities, and we hope that they can serve as a new baseline and benchmark for future work.

cs.CV↗

arXiv2025

SafeMVDrive: Multi-view Safety-Critical Driving Video Synthesis in the Real World Domain

Safety-critical scenarios are rare yet pivotal for evaluating and enhancing the robustness of autonomous driving systems. While existing methods generate safety-critical driving trajectories, simulations, or single-view videos, they fall short of meeting the demands of advanced end-to-end autonomous systems (E2E AD), which require real-world, multi-view video data. To bridge this gap, we introduce SafeMVDrive, the first framework designed to generate high-quality, safety-critical, multi-view driving videos grounded in real-world domains. SafeMVDrive strategically integrates a safety-critical trajectory generator with an advanced multi-view video generator. To tackle the challenges inherent in this integration, we first enhance scene understanding ability of the trajectory generator by incorporating visual context -- which is previously unavailable to such generator -- and leveraging a GRPO-finetuned vision-language model to achieve more realistic and context-aware trajectory generation. Second, recognizing that existing multi-view video generators struggle to render realistic collision events, we introduce a two-stage, controllable trajectory generation mechanism that produces collision-evasion trajectories, ensuring both video quality and safety-critical fidelity. Finally, we employ a diffusion-based multi-view video generator to synthesize high-quality safety-critical driving videos from the generated trajectories. Experiments conducted on an E2E AD planner demonstrate a significant increase in collision rate when tested with our generated data, validating the effectiveness of SafeMVDrive in stress-testing planning modules. Our code, examples, and datasets are publicly available at: https://zhoujiawei3.github.io/SafeMVDrive/.

cs.CV↗

arXiv2025

DeCLIP: Decoupled Learning for Open-Vocabulary Dense Perception

Dense visual prediction tasks have been constrained by their reliance on predefined categories, limiting their applicability in real-world scenarios where visual concepts are unbounded. While Vision-Language Models (VLMs) like CLIP have shown promise in open-vocabulary tasks, their direct application to dense prediction often leads to suboptimal performance due to limitations in local feature representation. In this work, we present our observation that CLIP's image tokens struggle to effectively aggregate information from spatially or semantically related regions, resulting in features that lack local discriminability and spatial consistency. To address this issue, we propose DeCLIP, a novel framework that enhances CLIP by decoupling the self-attention module to obtain ``content'' and ``context'' features respectively. The ``content'' features are aligned with image crop representations to improve local discriminability, while ``context'' features learn to retain the spatial correlations under the guidance of vision foundation models, such as DINO. Extensive experiments demonstrate that DeCLIP significantly outperforms existing methods across multiple open-vocabulary dense prediction tasks, including object detection and semantic segmentation. Code is available at \textcolor{magenta}{https://github.com/xiaomoguhz/DeCLIP}.

cs.CV↗

arXiv2025

From Mapping to Composing: A Two-Stage Framework for Zero-shot Composed Image Retrieval

Composed Image Retrieval (CIR) is a challenging multimodal task that retrieves a target image based on a reference image and accompanying modification text. Due to the high cost of annotating CIR triplet datasets, zero-shot (ZS) CIR has gained traction as a promising alternative. Existing studies mainly focus on projection-based methods, which map an image to a single pseudo-word token. However, these methods face three critical challenges: (1) insufficient pseudo-word token representation capacity, (2) discrepancies between training and inference phases, and (3) reliance on large-scale synthetic data. To address these issues, we propose a two-stage framework where the training is accomplished from mapping to composing. In the first stage, we enhance image-to-pseudo-word token learning by introducing a visual semantic injection module and a soft text alignment objective, enabling the token to capture richer and fine-grained image information. In the second stage, we optimize the text encoder using a small amount of synthetic triplet data, enabling it to effectively extract compositional semantics by combining pseudo-word tokens with modification text for accurate target image retrieval. The strong visual-to-pseudo mapping established in the first stage provides a solid foundation for the second stage, making our approach compatible with both high- and low-quality synthetic data, and capable of achieving significant performance gains with only a small amount of synthetic data. Extensive experiments were conducted on three public datasets, achieving superior performance compared to existing approaches.

cs.CV↗

arXiv2024

Referencing Where to Focus: Improving VisualGrounding with Referential Query

Visual Grounding aims to localize the referring object in an image given a natural language expression. Recent advancements in DETR-based visual grounding methods have attracted considerable attention, as they directly predict the coordinates of the target object without relying on additional efforts, such as pre-generated proposal candidates or pre-defined anchor boxes. However, existing research primarily focuses on designing stronger multi-modal decoder, which typically generates learnable queries by random initialization or by using linguistic embeddings. This vanilla query generation approach inevitably increases the learning difficulty for the model, as it does not involve any target-related information at the beginning of decoding. Furthermore, they only use the deepest image feature during the query learning process, overlooking the importance of features from other levels. To address these issues, we propose a novel approach, called RefFormer. It consists of the query adaption module that can be seamlessly integrated into CLIP and generate the referential query to provide the prior context for decoder, along with a task-specific decoder. By incorporating the referential query into the decoder, we can effectively mitigate the learning difficulty of the decoder, and accurately concentrate on the target object. Additionally, our proposed query adaption module can also act as an adapter, preserving the rich knowledge within CLIP without the need to tune the parameters of the backbone network. Extensive experiments demonstrate the effectiveness and efficiency of our proposed method, outperforming state-of-the-art approaches on five visual grounding benchmarks.

cs.CV↗

arXiv2024

Typicalness-Aware Learning for Failure Detection

Deep neural networks (DNNs) often suffer from the overconfidence issue, where incorrect predictions are made with high confidence scores, hindering the applications in critical systems. In this paper, we propose a novel approach called Typicalness-Aware Learning (TAL) to address this issue and improve failure detection performance. We observe that, with the cross-entropy loss, model predictions are optimized to align with the corresponding labels via increasing logit magnitude or refining logit direction. However, regarding atypical samples, the image content and their labels may exhibit disparities. This discrepancy can lead to overfitting on atypical samples, ultimately resulting in the overconfidence issue that we aim to address. To tackle the problem, we have devised a metric that quantifies the typicalness of each sample, enabling the dynamic adjustment of the logit magnitude during the training process. By allowing atypical samples to be adequately fitted while preserving reliable logit direction, the problem of overconfidence can be mitigated. TAL has been extensively evaluated on benchmark datasets, and the results demonstrate its superiority over existing failure detection methods. Specifically, TAL achieves a more than 5% improvement on CIFAR100 in terms of the Area Under the Risk-Coverage Curve (AURC) compared to the state-of-the-art. Code is available at https://github.com/liuyijungoon/TAL.

cs.CV↗

arXiv2024

Decoupled Kullback-Leibler Divergence Loss

In this paper, we delve deeper into the Kullback-Leibler (KL) Divergence loss and mathematically prove that it is equivalent to the Decoupled Kullback-Leibler (DKL) Divergence loss that consists of 1) a weighted Mean Square Error (wMSE) loss and 2) a Cross-Entropy loss incorporating soft labels. Thanks to the decomposed formulation of DKL loss, we have identified two areas for improvement. Firstly, we address the limitation of KL/DKL in scenarios like knowledge distillation by breaking its asymmetric optimization property. This modification ensures that the $\mathbf{w}$MSE component is always effective during training, providing extra constructive cues. Secondly, we introduce class-wise global information into KL/DKL to mitigate bias from individual samples. With these two enhancements, we derive the Improved Kullback-Leibler (IKL) Divergence loss and evaluate its effectiveness by conducting experiments on CIFAR-10/100 and ImageNet datasets, focusing on adversarial training, and knowledge distillation tasks. The proposed approach achieves new state-of-the-art adversarial robustness on the public leaderboard -- RobustBench and competitive performance on knowledge distillation, demonstrating the substantial practical merits. Our code is available at https://github.com/jiequancui/DKL.

cs.CV↗

arXiv2024

Towards Large-scale 3D Representation Learning with Multi-dataset Point Prompt Training

The rapid advancement of deep learning models often attributes to their ability to leverage massive training data. In contrast, such privilege has not yet fully benefited 3D deep learning, mainly due to the limited availability of large-scale 3D datasets. Merging multiple available data sources and letting them collaboratively train a single model is a potential solution. However, due to the large domain gap between 3D point cloud datasets, such mixed supervision could adversely affect the model's performance and lead to degenerated performance (i.e., negative transfer) compared to single-dataset training. In view of this challenge, we introduce Point Prompt Training (PPT), a novel framework for multi-dataset synergistic learning in the context of 3D representation learning that supports multiple pre-training paradigms. Based on this framework, we propose Prompt-driven Normalization, which adapts the model to different datasets with domain-specific prompts and Language-guided Categorical Alignment that decently unifies the multiple-dataset label spaces by leveraging the relationship between label text. Extensive experiments verify that PPT can overcome the negative transfer associated with synergistic learning and produce generalizable representations. Notably, it achieves state-of-the-art performance on each dataset using a single weight-shared model with supervised multi-dataset training. Moreover, when served as a pre-training framework, it outperforms other pre-training approaches regarding representation quality and attains remarkable state-of-the-art performance across over ten diverse downstream tasks spanning both indoor and outdoor 3D scenarios.

cs.CV↗

arXiv2024

Explore the Potential of CLIP for Training-Free Open Vocabulary Semantic Segmentation

CLIP, as a vision-language model, has significantly advanced Open-Vocabulary Semantic Segmentation (OVSS) with its zero-shot capabilities. Despite its success, its application to OVSS faces challenges due to its initial image-level alignment training, which affects its performance in tasks requiring detailed local context. Our study delves into the impact of CLIP's [CLS] token on patch feature correlations, revealing a dominance of "global" patches that hinders local feature discrimination. To overcome this, we propose CLIPtrase, a novel training-free semantic segmentation strategy that enhances local feature awareness through recalibrated self-correlation among patches. This approach demonstrates notable improvements in segmentation accuracy and the ability to maintain semantic coherence across objects.Experiments show that we are 22.3% ahead of CLIP on average on 9 segmentation benchmarks, outperforming existing state-of-the-art training-free methods.The code are made publicly available at: https://github.com/leaves162/CLIPtrase.

cs.CV↗