Semantic querying in complex 3D scenes through free-form language presents a significant challenge. Existing 3D scene understanding methods use large-scale training data and CLIP to align text queries with 3D semantic features. However, their reliance on predefined vocabulary priors from training data hinders free-form semantic querying. Besides, recent advanced methods rely on LLMs for scene understanding but lack comprehensive 3D scene-level information and often overlook the potential inconsistencies in LLM-generated outputs. In our paper, we propose FreeQ-Graph, which enables Free-form Querying with a semantic consistent scene Graph for 3D scene understanding. The core idea is to encode free-form queries from a complete and accurate 3D scene graph without predefined vocabularies, and to align them with 3D consistent semantic labels, which accomplished through three key steps. We initiate by constructing a complete and accurate 3D scene graph that maps free-form objects and their relations through LLM and LVLM guidance, entirely free from training data or predefined priors. Most importantly, we align graph nodes with accurate semantic labels by leveraging 3D semantic aligned features from merged superpoints, enhancing 3D semantic consistency. To enable free-form semantic querying, we then design an LLM-based reasoning algorithm that combines scene-level and object-level information to intricate reasoning. We conducted extensive experiments on 3D semantic grounding, segmentation, and complex querying tasks, while also validating the accuracy of graph generation. Experiments on 6 datasets show that our model excels in both complex free-form semantic queries and intricate relational reasoning.
We introduce STSBench, a scenario-based framework to benchmark the holistic understanding of vision-language models (VLMs) for autonomous driving. The framework automatically mines pre-defined traffic scenarios from any dataset using ground-truth annotations, provides an intuitive user interface for efficient human verification, and generates multiple-choice questions for model evaluation. Applied to the NuScenes dataset, we present STSnu, the first benchmark that evaluates the spatio-temporal reasoning capabilities of VLMs based on comprehensive 3D perception. Existing benchmarks typically target off-the-shelf or fine-tuned VLMs for images or videos from a single viewpoint and focus on semantic tasks such as object recognition, dense captioning, risk assessment, or scene understanding. In contrast, STSnu evaluates driving expert VLMs for end-to-end driving, operating on videos from multi-view cameras or LiDAR. It specifically assesses their ability to reason about both ego-vehicle actions and complex interactions among traffic participants, a crucial capability for autonomous vehicles. The benchmark features 43 diverse scenarios spanning multiple views and frames, resulting in 971 human-verified multiple-choice questions. A thorough evaluation uncovers critical shortcomings in existing models' ability to reason about fundamental traffic dynamics in complex environments. These findings highlight the urgent need for architectural advances that explicitly model spatio-temporal reasoning. By addressing a core gap in spatio-temporal evaluation, STSBench enables the development of more robust and explainable VLMs for autonomous driving.
Large Vision Language Models (LVLMs) have shown strong capabilities in understanding and analyzing visual scenes across various domains. However, in the context of autonomous driving, their limited comprehension of 3D environments restricts their effectiveness in achieving a complete and safe understanding of dynamic surroundings. To address this, we introduce V3LMA, a novel approach that enhances 3D scene understanding by integrating Large Language Models (LLMs) with LVLMs. V3LMA leverages textual descriptions generated from object detections and video inputs, significantly boosting performance without requiring fine-tuning. Through a dedicated preprocessing pipeline that extracts 3D object data, our method improves situational awareness and decision-making in complex traffic scenarios, achieving a score of 0.56 on the LingoQA benchmark. We further explore different fusion strategies and token combinations with the goal of advancing the interpretation of traffic scenes, ultimately enabling safer autonomous driving systems.
Semantic Scene Completion (SSC) has emerged as a pivotal approach for jointly learning scene geometry and semantics, enabling downstream applications such as navigation in mobile robotics. The recent generalization to Panoptic Scene Completion (PSC) advances the SSC domain by integrating instance-level information, thereby enhancing object-level sensitivity in scene understanding. While PSC was introduced using LiDAR modality, methods based on camera images remain largely unexplored. Moreover, recent Transformer-based SSC approaches utilize a fixed set of learned queries to reconstruct objects within the scene volume. Although these queries are typically updated with image context during training, they remain static at test time, limiting their ability to dynamically adapt specifically to the observed scene. To overcome these limitations, we propose IPFormer, the first approach that leverages context-adaptive instance proposals at train and test time to address vision-based 3D Panoptic Scene Completion. Specifically, IPFormer adaptively initializes these queries as panoptic instance proposals derived from image context and further refines them through attention-based encoding and decoding to reason about semantic instance-voxel relationships. Experimental results show that our approach surpasses state-of-the-art methods in overall panoptic metrics PQ$^\dagger$ and PQ-All, matches performance in individual metrics, and achieves a runtime reduction exceeding 14$\times$. Furthermore, our ablation studies reveal that dynamically deriving instance proposals from image context, as opposed to random initialization, leads to a 3.62% increase in PQ-All and a remarkable average improvement of 18.65% in combined Thing-metrics. These results highlight our introduction of context-adaptive instance proposals as a pioneering effort in addressing vision-based 3D Panoptic Scene Completion.
Recent advances in text-to-3D scene generation have demonstrated significant potential to transform content creation across multiple industries. Although the research community has made impressive progress in addressing the challenges of this complex task, existing methods often generate environments that are only front-facing, lack visual fidelity, exhibit limited scene understanding, and are typically fine-tuned for either indoor or outdoor settings. In this work, we address these issues and propose DreamAnywhere, a modular system for the fast generation and prototyping of 3D scenes. Our system synthesizes a 360{\deg} panoramic image from text, decomposes it into background and objects, constructs a complete 3D representation through hybrid inpainting, and lifts object masks to detailed 3D objects that are placed in the virtual environment. DreamAnywhere supports immersive navigation and intuitive object-level editing, making it ideal for scene exploration, visual mock-ups, and rapid prototyping -- all with minimal manual modeling. These features make our system particularly suitable for low-budget movie production, enabling quick iteration on scene layout and visual tone without the overhead of traditional 3D workflows. Our modular pipeline is highly customizable as it allows components to be replaced independently. Compared to current state-of-the-art text and image-based 3D scene generation approaches, DreamAnywhere shows significant improvements in coherence in novel view synthesis and achieves competitive image quality, demonstrating its effectiveness across diverse and challenging scenarios. A comprehensive user study demonstrates a clear preference for our method over existing approaches, validating both its technical robustness and practical usefulness.
Driving in safety-critical scenarios requires quick, context-aware decision-making grounded in both situational understanding and experiential reasoning. Large Language Models (LLMs), with their powerful general-purpose reasoning capabilities, offer a promising foundation for such decision-making. However, their direct application to autonomous driving remains limited due to challenges in domain adaptation, contextual grounding, and the lack of experiential knowledge needed to make reliable and interpretable decisions in dynamic, high-risk environments. To address this gap, this paper presents a Case-Based Reasoning Augmented Large Language Model (CBR-LLM) framework for evasive maneuver decision-making in complex risk scenarios. Our approach integrates semantic scene understanding from dashcam video inputs with the retrieval of relevant past driving cases, enabling LLMs to generate maneuver recommendations that are both context-sensitive and human-aligned. Experiments across multiple open-source LLMs show that our framework improves decision accuracy, justification quality, and alignment with human expert behavior. Risk-aware prompting strategies further enhance performance across diverse risk types, while similarity-based case retrieval consistently outperforms random sampling in guiding in-context learning. Case studies further demonstrate the framework's robustness in challenging real-world conditions, underscoring its potential as an adaptive and trustworthy decision-support tool for intelligent driving systems.
360 video captures the complete surrounding scenes with the ultra-large field of view of 360X180. This makes 360 scene understanding tasks, eg, segmentation and tracking, crucial for appications, such as autonomous driving, robotics. With the recent emergence of foundation models, the community is, however, impeded by the lack of large-scale, labelled real-world datasets. This is caused by the inherent spherical properties, eg, severe distortion in polar regions, and content discontinuities, rendering the annotation costly yet complex. This paper introduces Leader360V, the first large-scale, labeled real-world 360 video datasets for instance segmentation and tracking. Our datasets enjoy high scene diversity, ranging from indoor and urban settings to natural and dynamic outdoor scenes. To automate annotation, we design an automatic labeling pipeline, which subtly coordinates pre-trained 2D segmentors and large language models to facilitate the labeling. The pipeline operates in three novel stages. Specifically, in the Initial Annotation Phase, we introduce a Semantic- and Distortion-aware Refinement module, which combines object mask proposals from multiple 2D segmentors with LLM-verified semantic labels. These are then converted into mask prompts to guide SAM2 in generating distortion-aware masks for subsequent frames. In the Auto-Refine Annotation Phase, missing or incomplete regions are corrected either by applying the SDR again or resolving the discontinuities near the horizontal borders. The Manual Revision Phase finally incorporates LLMs and human annotators to further refine and validate the annotations. Extensive user studies and evaluations demonstrate the effectiveness of our labeling pipeline. Meanwhile, experiments confirm that Leader360V significantly enhances model performance for 360 video segmentation and tracking, paving the way for more scalable 360 scene understanding.
3D visual grounding (3DVG) is a critical task in scene understanding that aims to identify objects in 3D scenes based on text descriptions. However, existing methods rely on separately pre-trained vision and text encoders, resulting in a significant gap between the two modalities in terms of spatial geometry and semantic categories. This discrepancy often causes errors in object positioning and classification. The paper proposes UniSpace-3D, which innovatively introduces a unified representation space for 3DVG, effectively bridging the gap between visual and textual features. Specifically, UniSpace-3D incorporates three innovative designs: i) a unified representation encoder that leverages the pre-trained CLIP model to map visual and textual features into a unified representation space, effectively bridging the gap between the two modalities; ii) a multi-modal contrastive learning module that further reduces the modality gap; iii) a language-guided query selection module that utilizes the positional and semantic information to identify object candidate points aligned with textual descriptions. Extensive experiments demonstrate that UniSpace-3D outperforms baseline models by at least 2.24% on the ScanRefer and Nr3D/Sr3D datasets. The code will be made available upon acceptance of the paper.
Accurate prediction of pedestrian trajectories is essential for applications in robotics and surveillance systems. While existing approaches primarily focus on social interactions between pedestrians, they often overlook the rich environmental context that significantly shapes human movement patterns. In this paper, we propose SceneAware, a novel framework that explicitly incorporates scene understanding to enhance trajectory prediction accuracy. Our method leverages a Vision Transformer~(ViT) scene encoder to process environmental context from static scene images, while Multi-modal Large Language Models~(MLLMs) generate binary walkability masks that distinguish between accessible and restricted areas during training. We combine a Transformer-based trajectory encoder with the ViT-based scene encoder, capturing both temporal dynamics and spatial constraints. The framework integrates collision penalty mechanisms that discourage predicted trajectories from violating physical boundaries, ensuring physically plausible predictions. SceneAware is implemented in both deterministic and stochastic variants. Comprehensive experiments on the ETH/UCY benchmark datasets show that our approach outperforms state-of-the-art methods, with more than 50\% improvement over previous models. Our analysis based on different trajectory categories shows that the model performs consistently well across various types of pedestrian movement. This highlights the importance of using explicit scene information and shows that our scene-aware approach is both effective and reliable in generating accurate and physically plausible predictions. Code is available at: https://github.com/juho127/SceneAware.
The integration of Large Language Models (LLMs) with computer vision is profoundly transforming perception tasks like image segmentation. For intelligent transportation systems (ITS), where accurate scene understanding is critical for safety and efficiency, this new paradigm offers unprecedented capabilities. This survey systematically reviews the emerging field of LLM-augmented image segmentation, focusing on its applications, challenges, and future directions within ITS. We provide a taxonomy of current approaches based on their prompting mechanisms and core architectures, and we highlight how these innovations can enhance road scene understanding for autonomous driving, traffic monitoring, and infrastructure maintenance. Finally, we identify key challenges, including real-time performance and safety-critical reliability, and outline a perspective centered on explainable, human-centric AI as a prerequisite for the successful deployment of this technology in next-generation transportation systems.
Holistic 3D scene understanding, which jointly models geometry, appearance, and semantics, is crucial for applications like augmented reality and robotic interaction. Existing feed-forward 3D scene understanding methods (e.g., LSM) are limited to extracting language-based semantics from scenes, failing to achieve holistic scene comprehension. Additionally, they suffer from low-quality geometry reconstruction and noisy artifacts. In contrast, per-scene optimization methods rely on dense input views, which reduces practicality and increases complexity during deployment. In this paper, we propose SemanticSplat, a feed-forward semantic-aware 3D reconstruction method, which unifies 3D Gaussians with latent semantic attributes for joint geometry-appearance-semantics modeling. To predict the semantic anisotropic Gaussians, SemanticSplat fuses diverse feature fields (e.g., LSeg, SAM) with a cost volume representation that stores cross-view feature similarities, enhancing coherent and accurate scene comprehension. Leveraging a two-stage distillation framework, SemanticSplat reconstructs a holistic multi-modal semantic feature field from sparse-view images. Experiments demonstrate the effectiveness of our method for 3D scene understanding tasks like promptable and open-vocabulary segmentation. Video results are available at https://semanticsplat.github.io.
Generative models have gained significant attention in novel view synthesis (NVS) by alleviating the reliance on dense multi-view captures. However, existing methods typically fall into a conventional paradigm, where generative models first complete missing areas in 2D, followed by 3D recovery techniques to reconstruct the scene, which often results in overly smooth surfaces and distorted geometry, as generative models struggle to infer 3D structure solely from RGB data. In this paper, we propose SceneCompleter, a novel framework that achieves 3D-consistent generative novel view synthesis through dense 3D scene completion. SceneCompleter achieves both visual coherence and 3D-consistent generative scene completion through two key components: (1) a geometry-appearance dual-stream diffusion model that jointly synthesizes novel views in RGBD space; (2) a scene embedder that encodes a more holistic scene understanding from the reference image. By effectively fusing structural and textural information, our method demonstrates superior coherence and plausibility in generative novel view synthesis across diverse datasets. Project Page: https://chen-wl20.github.io/SceneCompleter
Image captioning involves generating textual descriptions from input images, bridging the gap between computer vision and natural language processing. Recent advancements in transformer-based models have significantly improved caption generation by leveraging attention mechanisms for better scene understanding. While various surveys have explored deep learning-based approaches for image captioning, few have comprehensively analyzed attention-based transformer models across multiple languages. This survey reviews attention-based image captioning models, categorizing them into transformer-based, deep learning-based, and hybrid approaches. It explores benchmark datasets, discusses evaluation metrics such as BLEU, METEOR, CIDEr, and ROUGE, and highlights challenges in multilingual captioning. Additionally, this paper identifies key limitations in current models, including semantic inconsistencies, data scarcity in non-English languages, and limitations in reasoning ability. Finally, we outline future research directions, such as multimodal learning, real-time applications in AI-powered assistants, healthcare, and forensic analysis. This survey serves as a comprehensive reference for researchers aiming to advance the field of attention-based image captioning.
SpatialLM is a large language model designed to process 3D point cloud data and generate structured 3D scene understanding outputs. These outputs include architectural elements like walls, doors, windows, and oriented object boxes with their semantic categories. Unlike previous methods which exploit task-specific network designs, our model adheres to the standard multimodal LLM architecture and is fine-tuned directly from open-source LLMs. To train SpatialLM, we collect a large-scale, high-quality synthetic dataset consisting of the point clouds of 12,328 indoor scenes (54,778 rooms) with ground-truth 3D annotations, and conduct a careful study on various modeling and training decisions. On public benchmarks, our model gives state-of-the-art performance in layout estimation and competitive results in 3D object detection. With that, we show a feasible path for enhancing the spatial understanding capabilities of modern LLMs for applications in augmented reality, embodied robotics, and more.
Visible images offer rich texture details, while infrared images emphasize salient targets. Fusing these complementary modalities enhances scene understanding, particularly for advanced vision tasks under challenging conditions. Recently, deep learning-based fusion methods have gained attention, but current evaluations primarily rely on general-purpose metrics without standardized benchmarks or downstream task performance. Additionally, the lack of well-developed dual-spectrum datasets and fair algorithm comparisons hinders progress. To address these gaps, we construct a high-quality dual-spectrum dataset captured in campus environments, comprising 1,369 well-aligned visible-infrared image pairs across four representative scenarios: daytime, nighttime, smoke occlusion, and underpasses. We also propose a comprehensive and fair evaluation framework that integrates fusion speed, general metrics, and object detection performance using the lang-segment-anything model to ensure fairness in downstream evaluation. Extensive experiments benchmark several state-of-the-art fusion algorithms under this framework. Results demonstrate that fusion models optimized for downstream tasks achieve superior performance in target detection, especially in low-light and occluded scenes. Notably, some algorithms that perform well on general metrics do not translate to strong downstream performance, highlighting limitations of current evaluation practices and validating the necessity of our proposed framework. The main contributions of this work are: (1)a campus-oriented dual-spectrum dataset with diverse and challenging scenes; (2) a task-aware, comprehensive evaluation framework; and (3) thorough comparative analysis of leading fusion methods across multiple datasets, offering insights for future development.
3D Gaussian Splatting (3DGS) has emerged as a powerful representation for neural scene reconstruction, offering high-quality novel view synthesis while maintaining computational efficiency. In this paper, we extend the capabilities of 3DGS beyond pure scene representation by introducing an approach for open-vocabulary 3D instance segmentation without requiring manual labeling, termed OpenSplat3D. Our method leverages feature-splatting techniques to associate semantic information with individual Gaussians, enabling fine-grained scene understanding. We incorporate Segment Anything Model instance masks with a contrastive loss formulation as guidance for the instance features to achieve accurate instance-level segmentation. Furthermore, we utilize language embeddings of a vision-language model, allowing for flexible, text-driven instance identification. This combination enables our system to identify and segment arbitrary objects in 3D scenes based on natural language descriptions. We show results on LERF-mask and LERF-OVS as well as the full ScanNet++ validation set, demonstrating the effectiveness of our approach.
In recent years, numerous tasks have been proposed to encourage model to develop specified capability in understanding audio-visual scene, primarily categorized into temporal localization, spatial localization, spatio-temporal reasoning, and pixel-level understanding. Instead, human possesses a unified understanding ability for diversified tasks. Therefore, designing an audio-visual model with general capability to unify these tasks is of great value. However, simply joint training for all tasks can lead to interference due to the heterogeneity of audiovisual data and complex relationship among tasks. We argue that this problem can be solved through explicit cooperation among tasks. To achieve this goal, we propose a unified learning method which achieves explicit inter-task cooperation from both the perspectives of data and model thoroughly. Specifically, considering the labels of existing datasets are simple words, we carefully refine these datasets and construct an Audio-Visual Unified Instruction-tuning dataset with Explicit reasoning process (AV-UIE), which clarifies the cooperative relationship among tasks. Subsequently, to facilitate concrete cooperation in learning stage, an interaction-aware LoRA structure with multiple LoRA heads is designed to learn different aspects of audiovisual data interaction. By unifying the explicit cooperation across the data and model aspect, our method not only surpasses existing unified audio-visual model on multiple tasks, but also outperforms most specialized models for certain tasks. Furthermore, we also visualize the process of explicit cooperation and surprisingly find that each LoRA head has certain audio-visual understanding ability. Code and dataset: https://github.com/GeWu-Lab/Crab
3D Gaussian Splatting (3DGS) serves as a highly performant and efficient encoding of scene geometry, appearance, and semantics. Moreover, grounding language in 3D scenes has proven to be an effective strategy for 3D scene understanding. Current Language Gaussian Splatting line of work fall into three main groups: (i) per-scene optimization-based, (ii) per-scene optimization-free, and (iii) generalizable approach. However, most of them are evaluated only on rendered 2D views of a handful of scenes and viewpoints close to the training views, limiting ability and insight into holistic 3D understanding. To address this gap, we propose the first large-scale benchmark that systematically assesses these three groups of methods directly in 3D space, evaluating on 1060 scenes across three indoor datasets and one outdoor dataset. Benchmark results demonstrate a clear advantage of the generalizable paradigm, particularly in relaxing the scene-specific limitation, enabling fast feed-forward inference on novel scenes, and achieving superior segmentation performance. We further introduce GaussianWorld-49K a carefully curated 3DGS dataset comprising around 49K diverse indoor and outdoor scenes obtained from multiple sources, with which we demonstrate the generalizable approach could harness strong data priors. Our codes, benchmark, and datasets will be made public to accelerate research in generalizable 3DGS scene understanding.
While vision-language models (VLMs) have demonstrated promising capabilities in reasoning and planning for embodied agents, their ability to comprehend physical phenomena, particularly within structured 3D environments, remains severely limited. To close this gap, we introduce PhyBlock, a progressive benchmark designed to assess VLMs on physical understanding and planning through robotic 3D block assembly tasks. PhyBlock integrates a novel four-level cognitive hierarchy assembly task alongside targeted Visual Question Answering (VQA) samples, collectively aimed at evaluating progressive spatial reasoning and fundamental physical comprehension, including object properties, spatial relationships, and holistic scene understanding. PhyBlock includes 2600 block tasks (400 assembly tasks, 2200 VQA tasks) and evaluates models across three key dimensions: partial completion, failure diagnosis, and planning robustness. We benchmark 21 state-of-the-art VLMs, highlighting their strengths and limitations in physically grounded, multi-step planning. Our empirical findings indicate that the performance of VLMs exhibits pronounced limitations in high-level planning and reasoning capabilities, leading to a notable decline in performance for the growing complexity of the tasks. Error analysis reveals persistent difficulties in spatial orientation and dependency reasoning. Surprisingly, chain-of-thought prompting offers minimal improvements, suggesting spatial tasks heavily rely on intuitive model comprehension. We position PhyBlock as a unified testbed to advance embodied reasoning, bridging vision-language understanding and real-world physical problem-solving.
Visual localization remains challenging in dynamic environments where fluctuating lighting, adverse weather, and moving objects disrupt appearance cues. Despite advances in feature representation, current absolute pose regression methods struggle to maintain consistency under varying conditions. To address this challenge, we propose a framework that synergistically combines multi-scale feature learning with semantic scene understanding. Our approach employs a hierarchical Transformer with cross-scale attention to fuse geometric details and contextual cues, preserving spatial precision while adapting to environmental changes. We improve the performance of this architecture with semantic supervision via neural scene representation during training, guiding the network to learn view-invariant features that encode persistent structural information while suppressing complex environmental interference. Experiments on TartanAir demonstrate that our approach outperforms existing pose regression methods in challenging scenarios with dynamic objects, illumination changes, and occlusions. Our findings show that integrating multi-scale processing with semantic guidance offers a promising strategy for robust visual localization in real-world dynamic environments.