Understanding complex scenes at multiple levels of abstraction remains a formidable challenge in computer vision. To address this, we introduce Nested Neural Feature Fields (N2F2), a novel approach that employs hierarchical supervision to learn a single feature field, wherein different dimensions within the same high-dimensional feature encode scene properties at varying granularities. Our method allows for a flexible definition of hierarchies, tailored to either the physical dimensions or semantics or both, thereby enabling a comprehensive and nuanced understanding of scenes. We leverage a 2D class-agnostic segmentation model to provide semantically meaningful pixel groupings at arbitrary scales in the image space, and query the CLIP vision-encoder to obtain language-aligned embeddings for each of these segments. Our proposed hierarchical supervision method then assigns different nested dimensions of the feature field to distill the CLIP embeddings using deferred volumetric rendering at varying physical scales, creating a coarse-to-fine representation. Extensive experiments show that our approach outperforms the state-of-the-art feature field distillation methods on tasks such as open-vocabulary 3D segmentation and localization, demonstrating the effectiveness of the learned nested feature field.
Multi-class multi-instance segmentation is the task of identifying masks for multiple object classes and multiple instances of the same class within an image. The foundational Segment Anything Model (SAM) is designed for promptable multi-class multi-instance segmentation but tends to output part or sub-part masks in the "everything" mode for various real-world applications. Whole object segmentation masks play a crucial role for indoor scene understanding, especially in robotics applications. We propose a new domain invariant Real-to-Simulation (Real-Sim) fine-tuning strategy for SAM. We use object images and ground truth data collected from Ai2Thor simulator during fine-tuning (real-to-sim). To allow our Segment Any Object Model (SAOM) to work in the "everything" mode, we propose the novel nearest neighbour assignment method, updating point embeddings for each ground-truth mask. SAOM is evaluated on our own dataset collected from Ai2Thor simulator. SAOM significantly improves on SAM, with a 28% increase in mIoU and a 25% increase in mAcc for 54 frequently-seen indoor object classes. Moreover, our Real-to-Simulation fine-tuning strategy demonstrates promising generalization performance in real environments without being trained on the real-world data (sim-to-real). The dataset and the code will be released after publication.
The ability to understand and reason the 3D real world is a crucial milestone towards artificial general intelligence. The current common practice is to finetune Large Language Models (LLMs) with 3D data and texts to enable 3D understanding. Despite their effectiveness, these approaches are inherently limited by the scale and diversity of the available 3D data. Alternatively, in this work, we introduce Agent3D-Zero, an innovative 3D-aware agent framework addressing the 3D scene understanding in a zero-shot manner. The essence of our approach centers on reconceptualizing the challenge of 3D scene perception as a process of understanding and synthesizing insights from multiple images, inspired by how our human beings attempt to understand 3D scenes. By consolidating this idea, we propose a novel way to make use of a Large Visual Language Model (VLM) via actively selecting and analyzing a series of viewpoints for 3D understanding. Specifically, given an input 3D scene, Agent3D-Zero first processes a bird's-eye view image with custom-designed visual prompts, then iteratively chooses the next viewpoints to observe and summarize the underlying knowledge. A distinctive advantage of Agent3D-Zero is the introduction of novel visual prompts, which significantly unleash the VLMs' ability to identify the most informative viewpoints and thus facilitate observing 3D scenes. Extensive experiments demonstrate the effectiveness of the proposed framework in understanding diverse and previously unseen 3D environments.
Human-centered dynamic scene understanding plays a pivotal role in enhancing the capability of robotic and autonomous systems, in which Video-based Human-Object Interaction (V-HOI) detection is a crucial task in semantic scene understanding, aimed at comprehensively understanding HOI relationships within a video to benefit the behavioral decisions of mobile robots and autonomous driving systems. Although previous V-HOI detection models have made significant strides in accurate detection on specific datasets, they still lack the general reasoning ability like human beings to effectively induce HOI relationships. In this study, we propose V-HOI Multi-LLMs Collaborated Reasoning (V-HOI MLCR), a novel framework consisting of a series of plug-and-play modules that could facilitate the performance of current V-HOI detection models by leveraging the strong reasoning ability of different off-the-shelf pre-trained large language models (LLMs). We design a two-stage collaboration system of different LLMs for the V-HOI task. Specifically, in the first stage, we design a Cross-Agents Reasoning scheme to leverage the LLM conduct reasoning from different aspects. In the second stage, we perform Multi-LLMs Debate to get the final reasoning answer based on the different knowledge in different LLMs. Additionally, we devise an auxiliary training strategy that utilizes CLIP, a large vision-language model to enhance the base V-HOI models' discriminative ability to better cooperate with LLMs. We validate the superiority of our design by demonstrating its effectiveness in improving the prediction accuracy of the base V-HOI model via reasoning from multiple perspectives.
Self-supervised 3D representation learning aims to learn effective representations from large-scale unlabeled point clouds. Most existing approaches adopt point discrimination as the pretext task, which assigns matched points in two distinct views as positive pairs and unmatched points as negative pairs. However, this approach often results in semantically identical points having dissimilar representations, leading to a high number of false negatives and introducing a "semantic conflict" problem. To address this issue, we propose GroupContrast, a novel approach that combines segment grouping and semantic-aware contrastive learning. Segment grouping partitions points into semantically meaningful regions, which enhances semantic coherence and provides semantic guidance for the subsequent contrastive representation learning. Semantic-aware contrastive learning augments the semantic information extracted from segment grouping and helps to alleviate the issue of "semantic conflict". We conducted extensive experiments on multiple 3D scene understanding tasks. The results demonstrate that GroupContrast learns semantically meaningful representations and achieves promising transfer learning performance.
Robots require a semantic understanding of their surroundings to operate in an efficient and explainable way in human environments. In the literature, there has been an extensive focus on object labeling and exhaustive scene graph generation; less effort has been focused on the task of purely identifying and mapping large semantic regions. The present work proposes a method for semantic region mapping via embodied navigation in indoor environments, generating a high-level representation of the knowledge of the agent. To enable region identification, the method uses a vision-to-language model to provide scene information for mapping. By projecting egocentric scene understanding into the global frame, the proposed method generates a semantic map as a distribution over possible region labels at each location. This mapping procedure is paired with a trained navigation policy to enable autonomous map generation. The proposed method significantly outperforms a variety of baselines, including an object-based system and a pretrained scene classifier, in experiments in a photorealistic simulator.
The rise of large language models (LLMs) and instruction tuning has led to the current trend of instruction-tuned large language and vision models (LLVMs). This trend involves either meticulously curating numerous instruction tuning datasets tailored to specific objectives or enlarging LLVMs to manage vast amounts of vision language (VL) data. However, current LLVMs have disregarded the detailed and comprehensive real-world scene understanding available from specialized computer vision (CV) models in visual perception tasks such as segmentation, detection, scene graph generation (SGG), and optical character recognition (OCR). Instead, the existing LLVMs rely mainly on the large capacity and emergent capabilities of their LLM backbones. Therefore, we present a new LLVM, Mixture of All Intelligence (MoAI), which leverages auxiliary visual information obtained from the outputs of external segmentation, detection, SGG, and OCR models. MoAI operates through two newly introduced modules: MoAI-Compressor and MoAI-Mixer. After verbalizing the outputs of the external CV models, the MoAI-Compressor aligns and condenses them to efficiently use relevant auxiliary visual information for VL tasks. MoAI-Mixer then blends three types of intelligence (1) visual features, (2) auxiliary features from the external CV models, and (3) language features by utilizing the concept of Mixture of Experts. Through this integration, MoAI significantly outperforms both open-source and closed-source LLVMs in numerous zero-shot VL tasks, particularly those related to real-world scene understanding such as object existence, positions, relations, and OCR without enlarging the model size or curating extra visual instruction tuning datasets.
Purpose: Advances in deep learning have resulted in effective models for surgical video analysis; however, these models often fail to generalize across medical centers due to domain shift caused by variations in surgical workflow, camera setups, and patient demographics. Recently, object-centric learning has emerged as a promising approach for improved surgical scene understanding, capturing and disentangling visual and semantic properties of surgical tools and anatomy to improve downstream task performance. In this work, we conduct a multi-centric performance benchmark of object-centric approaches, focusing on Critical View of Safety assessment in laparoscopic cholecystectomy, then propose an improved approach for unseen domain generalization. Methods: We evaluate four object-centric approaches for domain generalization, establishing baseline performance. Next, leveraging the disentangled nature of object-centric representations, we dissect one of these methods through a series of ablations (e.g. ignoring either visual or semantic features for downstream classification). Finally, based on the results of these ablations, we develop an optimized method specifically tailored for domain generalization, LG-DG, that includes a novel disentanglement loss function. Results: Our optimized approach, LG-DG, achieves an improvement of 9.28% over the best baseline approach. More broadly, we show that object-centric approaches are highly effective for domain generalization thanks to their modular approach to representation learning. Conclusion: We investigate the use of object-centric methods for unseen domain generalization, identify method-agnostic factors critical for performance, and present an optimized approach that substantially outperforms existing methods.
Monocular depth estimation is a crucial task in computer vision. While existing methods have shown impressive results under standard conditions, they often face challenges in reliably performing in scenarios such as low-light or rainy conditions due to the absence of diverse training data. This paper introduces a novel approach named Stealing Stable Diffusion (SSD) prior for robust monocular depth estimation. The approach addresses this limitation by utilizing stable diffusion to generate synthetic images that mimic challenging conditions. Additionally, a self-training mechanism is introduced to enhance the model's depth estimation capability in such challenging environments. To enhance the utilization of the stable diffusion prior further, the DINOv2 encoder is integrated into the depth model architecture, enabling the model to leverage rich semantic priors and improve its scene understanding. Furthermore, a teacher loss is introduced to guide the student models in acquiring meaningful knowledge independently, thus reducing their dependency on the teacher models. The effectiveness of the approach is evaluated on nuScenes and Oxford RobotCar, two challenging public datasets, with the results showing the efficacy of the method. Source code and weights are available at: https://github.com/hitcslj/SSD.
Embodied scene understanding serves as the cornerstone for autonomous agents to perceive, interpret, and respond to open driving scenarios. Such understanding is typically founded upon Vision-Language Models (VLMs). Nevertheless, existing VLMs are restricted to the 2D domain, devoid of spatial awareness and long-horizon extrapolation proficiencies. We revisit the key aspects of autonomous driving and formulate appropriate rubrics. Hereby, we introduce the Embodied Language Model (ELM), a comprehensive framework tailored for agents' understanding of driving scenes with large spatial and temporal spans. ELM incorporates space-aware pre-training to endow the agent with robust spatial localization capabilities. Besides, the model employs time-aware token selection to accurately inquire about temporal cues. We instantiate ELM on the reformulated multi-faced benchmark, and it surpasses previous state-of-the-art approaches in all aspects. All code, data, and models will be publicly shared.
Rapid and reliable identification of dynamic scene parts, also known as motion segmentation, is a key challenge for mobile sensors. Contemporary RGB camera-based methods rely on modeling camera and scene properties however, are often under-constrained and fall short in unknown categories. Event cameras have the potential to overcome these limitations, but corresponding methods have only been demonstrated in smaller-scale indoor environments with simplified dynamic objects. This work presents an event-based method for class-agnostic motion segmentation that can successfully be deployed across complex large-scale outdoor environments too. To this end, we introduce a novel divide-and-conquer pipeline that combines: (a) ego-motion compensated events, computed via a scene understanding module that predicts monocular depth and camera pose as auxiliary tasks, and (b) optical flow from a dedicated optical flow module. These intermediate representations are then fed into a segmentation module that predicts motion segmentation masks. A novel transformer-based temporal attention module in the segmentation module builds correlations across adjacent 'frames' to get temporally consistent segmentation masks. Our method sets the new state-of-the-art on the classic EV-IMO benchmark (indoors), where we achieve improvements of 2.19 moving object IoU (2.22 mIoU) and 4.52 point IoU respectively, as well as on a newly-generated motion segmentation and tracking benchmark (outdoors) based on the DSEC event dataset, termed DSEC-MOTS, where we show improvement of 12.91 moving object IoU.
Utilizing multi-view inputs to synthesize novel-view images, Neural Radiance Fields (NeRF) have emerged as a popular research topic in 3D vision. In this work, we introduce a Generalizable Semantic Neural Radiance Field (GSNeRF), which uniquely takes image semantics into the synthesis process so that both novel view images and the associated semantic maps can be produced for unseen scenes. Our GSNeRF is composed of two stages: Semantic Geo-Reasoning and Depth-Guided Visual rendering. The former is able to observe multi-view image inputs to extract semantic and geometry features from a scene. Guided by the resulting image geometry information, the latter performs both image and semantic rendering with improved performances. Our experiments not only confirm that GSNeRF performs favorably against prior works on both novel-view image and semantic segmentation synthesis but the effectiveness of our sampling strategy for visual rendering is further verified.
Human-centric 3D scene understanding has recently drawn increasing attention, driven by its critical impact on robotics. However, human-centric real-life scenarios are extremely diverse and complicated, and humans have intricate motions and interactions. With limited labeled data, supervised methods are difficult to generalize to general scenarios, hindering real-life applications. Mimicking human intelligence, we propose an unsupervised 3D detection method for human-centric scenarios by transferring the knowledge from synthetic human instances to real scenes. To bridge the gap between the distinct data representations and feature distributions of synthetic models and real point clouds, we introduce novel modules for effective instance-to-scene representation transfer and synthetic-to-real feature alignment. Remarkably, our method exhibits superior performance compared to current state-of-the-art techniques, achieving 87.8% improvement in mAP and closely approaching the performance of fully supervised methods (62.15 mAP vs. 69.02 mAP) on HuCenLife Dataset.
In the realm of computer vision, the integration of advanced techniques into the processing of RGB-D camera inputs poses a significant challenge, given the inherent complexities arising from diverse environmental conditions and varying object appearances. Therefore, this paper introduces FusionVision, an exhaustive pipeline adapted for the robust 3D segmentation of objects in RGB-D imagery. Traditional computer vision systems face limitations in simultaneously capturing precise object boundaries and achieving high-precision object detection on depth map as they are mainly proposed for RGB cameras. To address this challenge, FusionVision adopts an integrated approach by merging state-of-the-art object detection techniques, with advanced instance segmentation methods. The integration of these components enables a holistic (unified analysis of information obtained from both color \textit{RGB} and depth \textit{D} channels) interpretation of RGB-D data, facilitating the extraction of comprehensive and accurate object information. The proposed FusionVision pipeline employs YOLO for identifying objects within the RGB image domain. Subsequently, FastSAM, an innovative semantic segmentation model, is applied to delineate object boundaries, yielding refined segmentation masks. The synergy between these components and their integration into 3D scene understanding ensures a cohesive fusion of object detection and segmentation, enhancing overall precision in 3D object segmentation. The code and pre-trained models are publicly available at https://github.com/safouaneelg/FusionVision/.
Understanding a surgical scene is crucial for computer-assisted surgery systems to provide any intelligent assistance functionality. One way of achieving this scene understanding is via scene segmentation, where every pixel of a frame is classified and therefore identifies the visible structures and tissues. Progress on fully segmenting surgical scenes has been made using machine learning. However, such models require large amounts of annotated training data, containing examples of all relevant object classes. Such fully annotated datasets are hard to create, as every pixel in a frame needs to be annotated by medical experts and, therefore, are rarely available. In this work, we propose a method to combine multiple partially annotated datasets, which provide complementary annotations, into one model, enabling better scene segmentation and the use of multiple readily available datasets. Our method aims to combine available data with complementary labels by leveraging mutual exclusive properties to maximize information. Specifically, we propose to use positive annotations of other classes as negative samples and to exclude background pixels of binary annotations, as we cannot tell if they contain a class not annotated but predicted by the model. We evaluate our method by training a DeepLabV3 on the publicly available Dresden Surgical Anatomy Dataset, which provides multiple subsets of binary segmented anatomical structures. Our approach successfully combines 6 classes into one model, increasing the overall Dice Score by 4.4% compared to an ensemble of models trained on the classes individually. By including information on multiple classes, we were able to reduce confusion between stomach and colon by 24%. Our results demonstrate the feasibility of training a model on multiple datasets. This paves the way for future work further alleviating the need for one large, fully segmented datasets.
End-to-end models capable of handling multiple sub-tasks in parallel have become a new trend, thereby presenting significant challenges and opportunities for the integration of multiple tasks within the domain of 3D vision. The limitations of 3D data acquisition conditions have not only restricted the exploration of many innovative research problems but have also caused existing 3D datasets to predominantly focus on single tasks. This has resulted in a lack of systematic approaches and theoretical frameworks for 3D multi-task learning, with most efforts merely serving as auxiliary support to the primary task. In this paper, we introduce WHU-Synthetic, a large-scale 3D synthetic perception dataset designed for multi-task learning, from the initial data augmentation (upsampling and depth completion), through scene understanding (segmentation), to macro-level tasks (place recognition and 3D reconstruction). Collected in the same environmental domain, we ensure inherent alignment across sub-tasks to construct multi-task models without separate training methods. Besides, we implement several novel settings, making it possible to realize certain ideas that are difficult to achieve in real-world scenarios. This supports more adaptive and robust multi-task perception tasks, such as sampling on city-level models, providing point clouds with different densities, and simulating temporal changes. Using our dataset, we conduct several experiments to investigate mutual benefits between sub-tasks, revealing new observations, challenges, and opportunities for future research. The dataset is accessible at https://github.com/WHU-USI3DV/WHU-Synthetic.
Event cameras can record scene dynamics with high temporal resolution, providing rich scene details for monocular depth estimation (MDE) even at low-level illumination. Therefore, existing complementary learning approaches for MDE fuse intensity information from images and scene details from event data for better scene understanding. However, most methods directly fuse two modalities at pixel level, ignoring that the attractive complementarity mainly impacts high-level patterns that only occupy a few pixels. For example, event data is likely to complement contours of scene objects. In this paper, we discretize the scene into a set of high-level patterns to explore the complementarity and propose a Pattern-based Complementary learning architecture for monocular Depth estimation (PCDepth). Concretely, PCDepth comprises two primary components: a complementary visual representation learning module for discretizing the scene into high-level patterns and integrating complementary patterns across modalities and a refined depth estimator aimed at scene reconstruction and depth prediction while maintaining an efficiency-accuracy balance. Through pattern-based complementary learning, PCDepth fully exploits two modalities and achieves more accurate predictions than existing methods, especially in challenging nighttime scenarios. Extensive experiments on MVSEC and DSEC datasets verify the effectiveness and superiority of our PCDepth. Remarkably, compared with state-of-the-art, PCDepth achieves a 37.9% improvement in accuracy in MVSEC nighttime scenarios.
For human-centric large-scale scenes, fine-grained modeling for 3D human global pose and shape is significant for scene understanding and can benefit many real-world applications. In this paper, we present LiveHPS, a novel single-LiDAR-based approach for scene-level human pose and shape estimation without any limitation of light conditions and wearable devices. In particular, we design a distillation mechanism to mitigate the distribution-varying effect of LiDAR point clouds and exploit the temporal-spatial geometric and dynamic information existing in consecutive frames to solve the occlusion and noise disturbance. LiveHPS, with its efficient configuration and high-quality output, is well-suited for real-world applications. Moreover, we propose a huge human motion dataset, named FreeMotion, which is collected in various scenarios with diverse human poses, shapes and translations. It consists of multi-modal and multi-view acquisition data from calibrated and synchronized LiDARs, cameras, and IMUs. Extensive experiments on our new dataset and other public datasets demonstrate the SOTA performance and robustness of our approach. We will release our code and dataset soon.
The recent advancements in point cloud learning have enabled intelligent vehicles and robots to comprehend 3D environments better. However, processing large-scale 3D scenes remains a challenging problem, such that efficient downsampling methods play a crucial role in point cloud learning. Existing downsampling methods either require a huge computational burden or sacrifice fine-grained geometric information. For such purpose, this paper presents an advanced sampler that achieves both high accuracy and efficiency. The proposed method utilizes voxel centroid sampling as a foundation but effectively addresses the challenges regarding voxel size determination and the preservation of critical geometric cues. Specifically, we propose a Voxel Adaptation Module that adaptively adjusts voxel sizes with the reference of point-based downsampling ratio. This ensures that the sampling results exhibit a favorable distribution for comprehending various 3D objects or scenes. Meanwhile, we introduce a network compatible with arbitrary voxel sizes for sampling and feature extraction while maintaining high efficiency. The proposed approach is demonstrated with 3D object detection and 3D semantic segmentation. Compared to existing state-of-the-art methods, our approach achieves better accuracy on outdoor and indoor large-scale datasets, e.g. Waymo and ScanNet, with promising efficiency.
This report provides an overview of the challenge hosted at the OpenSUN3D Workshop on Open-Vocabulary 3D Scene Understanding held in conjunction with ICCV 2023. The goal of this workshop series is to provide a platform for exploration and discussion of open-vocabulary 3D scene understanding tasks, including but not limited to segmentation, detection and mapping. We provide an overview of the challenge hosted at the workshop, present the challenge dataset, the evaluation methodology, and brief descriptions of the winning methods. For additional details, please see https://opensun3d.github.io/index_iccv23.html.