Search

Recently, 3D Gaussian Splatting (3DGS) has revolutionized radiance field reconstruction, manifesting efficient and high-fidelity novel view synthesis. However, accurately representing surfaces, especially in large and complex scenarios, remains a significant challenge due to the unstructured nature of 3DGS. In this paper, we present CityGaussianV2, a novel approach for large-scale scene reconstruction that addresses critical challenges related to geometric accuracy and efficiency. Building on the favorable generalization capabilities of 2D Gaussian Splatting (2DGS), we address its convergence and scalability issues. Specifically, we implement a decomposed-gradient-based densification and depth regression technique to eliminate blurry artifacts and accelerate convergence. To scale up, we introduce an elongation filter that mitigates Gaussian count explosion caused by 2DGS degeneration. Furthermore, we optimize the CityGaussian pipeline for parallel training, achieving up to 10$\times$ compression, at least 25% savings in training time, and a 50% decrease in memory usage. We also established standard geometry benchmarks under large-scale scenes. Experimental results demonstrate that our method strikes a promising balance between visual quality, geometric accuracy, as well as storage and training costs. The project page is available at https://dekuliutesla.github.io/CityGaussianV2/., Comment: Project Page: https://dekuliutesla.github.io/CityGaussianV2/

In this paper, we propose OpenSatMap, a fine-grained, high-resolution satellite dataset for large-scale map construction. Map construction is one of the foundations of the transportation industry, such as navigation and autonomous driving. Extracting road structures from satellite images is an efficient way to construct large-scale maps. However, existing satellite datasets provide only coarse semantic-level labels with a relatively low resolution (up to level 19), impeding the advancement of this field. In contrast, the proposed OpenSatMap (1) has fine-grained instance-level annotations; (2) consists of high-resolution images (level 20); (3) is currently the largest one of its kind; (4) collects data with high diversity. Moreover, OpenSatMap covers and aligns with the popular nuScenes dataset and Argoverse 2 dataset to potentially advance autonomous driving technologies. By publishing and maintaining the dataset, we provide a high-quality benchmark for satellite-based map construction and downstream tasks like autonomous driving., Comment: NeurIPS 2024 D&B Track. Project Page:https://opensatmap.github.io/

Existing reconstruction-based novel view synthesis methods for driving scenes focus on synthesizing camera views along the recorded trajectory of the ego vehicle. Their image rendering performance will severely degrade on viewpoints falling out of the recorded trajectory, where camera rays are untrained. We propose FreeVS, a novel fully generative approach that can synthesize camera views on free new trajectories in real driving scenes. To control the generation results to be 3D consistent with the real scenes and accurate in viewpoint pose, we propose the pseudo-image representation of view priors to control the generation process. Viewpoint transformation simulation is applied on pseudo-images to simulate camera movement in each direction. Once trained, FreeVS can be applied to any validation sequences without reconstruction process and synthesis views on novel trajectories. Moreover, we propose two new challenging benchmarks tailored to driving scenes, which are novel camera synthesis and novel trajectory synthesis, emphasizing the freedom of viewpoints. Given that no ground truth images are available on novel trajectories, we also propose to evaluate the consistency of images synthesized on novel trajectories with 3D perception models. Experiments on the Waymo Open Dataset show that FreeVS has a strong image synthesis performance on both the recorded trajectories and novel trajectories. Project Page: https://freevs24.github.io/, Comment: Project Page: https://freevs24.github.io/

Enabling Large Language Models (LLMs) to handle a wider range of complex tasks (e.g., coding, math) has drawn great attention from many researchers. As LLMs continue to evolve, merely increasing the number of model parameters yields diminishing performance improvements and heavy computational costs. Recently, OpenAI's o1 model has shown that inference strategies (i.e., Test-time Compute methods) can also significantly enhance the reasoning capabilities of LLMs. However, the mechanisms behind these methods are still unexplored. In our work, to investigate the reasoning patterns of o1, we compare o1 with existing Test-time Compute methods (BoN, Step-wise BoN, Agent Workflow, and Self-Refine) by using OpenAI's GPT-4o as a backbone on general reasoning benchmarks in three domains (i.e., math, coding, commonsense reasoning). Specifically, first, our experiments show that the o1 model has achieved the best performance on most datasets. Second, as for the methods of searching diverse responses (e.g., BoN), we find the reward models' capability and the search space both limit the upper boundary of these methods. Third, as for the methods that break the problem into many sub-problems, the Agent Workflow has achieved better performance than Step-wise BoN due to the domain-specific system prompt for planning better reasoning processes. Fourth, it is worth mentioning that we have summarized six reasoning patterns of o1, and provided a detailed analysis on several reasoning benchmarks.

Large Language Models (LLMs) have displayed massive improvements in reasoning and decision-making skills and can hold natural conversations with users. Recently, many tool-use benchmark datasets have been proposed. However, existing datasets have the following limitations: (1). Insufficient evaluation scenarios (e.g., only cover limited tool-use scenes). (2). Extensive evaluation costs (e.g., GPT API costs). To address these limitations, in this work, we propose a multi-granularity tool-use benchmark for large language models called MTU-Bench. For the "multi-granularity" property, our MTU-Bench covers five tool usage scenes (i.e., single-turn and single-tool, single-turn and multiple-tool, multiple-turn and single-tool, multiple-turn and multiple-tool, and out-of-distribution tasks). Besides, all evaluation metrics of our MTU-Bench are based on the prediction results and the ground truth without using any GPT or human evaluation metrics. Moreover, our MTU-Bench is collected by transforming existing high-quality datasets to simulate real-world tool usage scenarios, and we also propose an instruction dataset called MTU-Instruct data to enhance the tool-use abilities of existing LLMs. Comprehensive experimental results demonstrate the effectiveness of our MTU-Bench. Code and data will be released at https: //github.com/MTU-Bench-Team/MTU-Bench.git.

Driving world models have gained increasing attention due to their ability to model complex physical dynamics. However, their superb modeling capability is yet to be fully unleashed due to the limited video diversity in current driving datasets. We introduce DrivingDojo, the first dataset tailor-made for training interactive world models with complex driving dynamics. Our dataset features video clips with a complete set of driving maneuvers, diverse multi-agent interplay, and rich open-world driving knowledge, laying a stepping stone for future world model development. We further define an action instruction following (AIF) benchmark for world models and demonstrate the superiority of the proposed dataset for generating action-controlled future predictions., Comment: Accepted to NeurIPS 2024. Project page: https://drivingdojo.github.io/

This paper introduces reconstructive visual instruction tuning (ROSS), a family of Large Multimodal Models (LMMs) that exploit vision-centric supervision signals. In contrast to conventional visual instruction tuning approaches that exclusively supervise text outputs, ROSS prompts LMMs to supervise visual outputs via reconstructing input images. By doing so, it capitalizes on the inherent richness and detail present within input images themselves, which are often lost in pure text supervision. However, producing meaningful feedback from natural images is challenging due to the heavy spatial redundancy of visual signals. To address this issue, ROSS employs a denoising objective to reconstruct latent representations of input images, avoiding directly regressing exact raw RGB values. This intrinsic activation design inherently encourages LMMs to maintain image detail, thereby enhancing their fine-grained comprehension capabilities and reducing hallucinations. Empirically, ROSS consistently brings significant improvements across different visual encoders and language models. In comparison with extrinsic assistance state-of-the-art alternatives that aggregate multiple visual experts, ROSS delivers competitive performance with a single SigLIP visual encoder, demonstrating the efficacy of our vision-centric supervision tailored for visual outputs.

In this paper, we introduce MIO, a novel foundation model built on multimodal tokens, capable of understanding and generating speech, text, images, and videos in an end-to-end, autoregressive manner. While the emergence of large language models (LLMs) and multimodal large language models (MM-LLMs) propels advancements in artificial general intelligence through their versatile capabilities, they still lack true any-to-any understanding and generation. Recently, the release of GPT-4o has showcased the remarkable potential of any-to-any LLMs for complex real-world tasks, enabling omnidirectional input and output across images, speech, and text. However, it is closed-source and does not support the generation of multimodal interleaved sequences. To address this gap, we present MIO, which is trained on a mixture of discrete tokens across four modalities using causal multimodal modeling. MIO undergoes a four-stage training process: (1) alignment pre-training, (2) interleaved pre-training, (3) speech-enhanced pre-training, and (4) comprehensive supervised fine-tuning on diverse textual, visual, and speech tasks. Our experimental results indicate that MIO exhibits competitive, and in some cases superior, performance compared to previous dual-modal baselines, any-to-any model baselines, and even modality-specific baselines. Moreover, MIO demonstrates advanced capabilities inherent to its any-to-any feature, such as interleaved video-text generation, chain-of-visual-thought reasoning, visual guideline generation, instructional image editing, etc., Comment: Technical Report. Codes and models are available in https://github.com/MIO-Team/MIO

In recent years, Large Language Models (LLMs) have demonstrated remarkable capabilities in various tasks (e.g., long-context understanding), and many benchmarks have been proposed. However, we observe that long text generation capabilities are not well investigated. Therefore, we introduce the Hierarchical Long Text Generation Benchmark (HelloBench), a comprehensive, in-the-wild, and open-ended benchmark to evaluate LLMs' performance in generating long text. Based on Bloom's Taxonomy, HelloBench categorizes long text generation tasks into five subtasks: open-ended QA, summarization, chat, text completion, and heuristic text generation. Besides, we propose Hierarchical Long Text Evaluation (HelloEval), a human-aligned evaluation method that significantly reduces the time and effort required for human evaluation while maintaining a high correlation with human evaluation. We have conducted extensive experiments across around 30 mainstream LLMs and observed that the current LLMs lack long text generation capabilities. Specifically, first, regardless of whether the instructions include explicit or implicit length constraints, we observe that most LLMs cannot generate text that is longer than 4000 words. Second, we observe that while some LLMs can generate longer text, many issues exist (e.g., severe repetition and quality degradation). Third, to demonstrate the effectiveness of HelloEval, we compare HelloEval with traditional metrics (e.g., ROUGE, BLEU, etc.) and LLM-as-a-Judge methods, which show that HelloEval has the highest correlation with human evaluation. We release our code in https://github.com/Quehry/HelloBench.

Recent advancements in multimodal large language models (MLLMs) have aimed to integrate and interpret data across diverse modalities. However, the capacity of these models to concurrently process and reason about multiple modalities remains inadequately explored, partly due to the lack of comprehensive modality-wise benchmarks. We introduce OmniBench, a novel benchmark designed to rigorously evaluate models' ability to recognize, interpret, and reason across visual, acoustic, and textual inputs simultaneously. We define models capable of such tri-modal processing as omni-language models (OLMs). OmniBench is distinguished by high-quality human annotations, ensuring that accurate responses require integrated understanding and reasoning across all three modalities. Our main findings reveal that: i) most OLMs exhibit critical limitations in instruction-following and reasoning capabilities within tri-modal contexts; and ii) most baselines models perform poorly (below 50\% accuracy) even when provided with alternative textual representations of images or/and audio. These results suggest that the ability to construct a consistent context from text, image, and audio is often overlooked in existing MLLM training paradigms. To address this gap, we curate an instruction tuning dataset of 84.5K training samples, OmniInstruct, for training OLMs to adapt to multimodal contexts. We advocate for future research to focus on developing more robust tri-modal integration techniques and training strategies to enhance OLM performance across diverse modalities. The codes and live leaderboard could be found at https://m-a-p.ai/OmniBench.

Foundation models like ChatGPT and Sora that are trained on a huge scale of data have made a revolutionary social impact. However, it is extremely challenging for sensors in many different fields to collect similar scales of natural images to train strong foundation models. To this end, this work presents a simple and effective framework SimMAT to study an open problem: the transferability from vision foundation models trained on natural RGB images to other image modalities of different physical properties (e.g., polarization). SimMAT consists of a modality-agnostic transfer layer (MAT) and a pretrained foundation model. We apply SimMAT to a representative vision foundation model Segment Anything Model (SAM) to support any evaluated new image modality. Given the absence of relevant benchmarks, we construct a new benchmark to evaluate the transfer learning performance. Our experiments confirm the intriguing potential of transferring vision foundation models in enhancing other sensors' performance. Specifically, SimMAT can improve the segmentation performance (mIoU) from 22.15% to 53.88% on average for evaluated modalities and consistently outperforms other baselines. We hope that SimMAT can raise awareness of cross-modal transfer learning and benefit various fields for better results with vision foundation models., Comment: Github link: https://github.com/mt-cly/SimMAT

Sound source localization aims to localize objects emitting the sound in visual scenes. Recent works obtaining impressive results typically rely on contrastive learning. However, the common practice of randomly sampling negatives in prior arts can lead to the false negative issue, where the sounds semantically similar to visual instance are sampled as negatives and incorrectly pushed away from the visual anchor/query. As a result, this misalignment of audio and visual features could yield inferior performance. To address this issue, we propose a novel audio-visual learning framework which is instantiated with two individual learning schemes: self-supervised predictive learning (SSPL) and semantic-aware contrastive learning (SACL). SSPL explores image-audio positive pairs alone to discover semantically coherent similarities between audio and visual features, while a predictive coding module for feature alignment is introduced to facilitate the positive-only learning. In this regard SSPL acts as a negative-free method to eliminate false negatives. By contrast, SACL is designed to compact visual features and remove false negatives, providing reliable visual anchor and audio negatives for contrast. Different from SSPL, SACL releases the potential of audio-visual contrastive learning, offering an effective alternative to achieve the same goal. Comprehensive experiments demonstrate the superiority of our approach over the state-of-the-arts. Furthermore, we highlight the versatility of the learned representation by extending the approach to audio-visual event classification and object detection tasks. Code and models are available at: https://github.com/zjsong/SACL., Comment: arXiv admin note: substantial text overlap with arXiv:2203.13412

Generating a realistic, large-scale 3D virtual city remains a complex challenge due to the involvement of numerous 3D assets, various city styles, and strict layout constraints. Existing approaches provide promising attempts at procedural content generation to create large-scale scenes using Blender agents. However, they face crucial issues such as difficulties in scaling up generation capability and achieving fine-grained control at the semantic layout level. To address these problems, we propose a novel multi-modal controllable procedural content generation method, named CityX, which enhances realistic, unbounded 3D city generation guided by multiple layout conditions, including OSM, semantic maps, and satellite images. Specifically, the proposed method contains a general protocol for integrating various PCG plugins and a multi-agent framework for transforming instructions into executable Blender actions. Through this effective framework, CityX shows the potential to build an innovative ecosystem for 3D scene generation by bridging the gap between the quality of generated assets and industrial requirements. Extensive experiments have demonstrated the effectiveness of our method in creating high-quality, diverse, and unbounded cities guided by multi-modal conditions. Our project page: https://cityx-lab.github.io.

The scarcity of large-scale 3D-text paired data poses a great challenge on open vocabulary 3D scene understanding, and hence it is popular to leverage internet-scale 2D data and transfer their open vocabulary capabilities to 3D models through knowledge distillation. However, the existing distillation-based 3D scene understanding approaches rely on the representation capacity of 2D models, disregarding the exploration of geometric priors and inherent representational advantages offered by 3D data. In this paper, we propose an effective approach, namely Geometry Guided Self-Distillation (GGSD), to learn superior 3D representations from 2D pre-trained models. Specifically, we first design a geometry guided distillation module to distill knowledge from 2D models, and then leverage the 3D geometric priors to alleviate the inherent noise in 2D models and enhance the representation learning process. Due to the advantages of 3D representation, the performance of the distilled 3D student model can significantly surpass that of the 2D teacher model. This motivates us to further leverage the representation advantages of 3D data through self-distillation. As a result, our proposed GGSD approach outperforms the existing open vocabulary 3D scene understanding methods by a large margin, as demonstrated by our experiments on both indoor and outdoor benchmark datasets.

3D object detection is an indispensable component for scene understanding. However, the annotation of large-scale 3D datasets requires significant human effort. To tackle this problem, many methods adopt weakly supervised 3D object detection that estimates 3D boxes by leveraging 2D boxes and scene/class-specific priors. However, these approaches generally depend on sophisticated manual priors, which is hard to generalize to novel categories and scenes. In this paper, we are motivated to propose a general approach, which can be easily adapted to new scenes and/or classes. A unified framework is developed for learning 3D object detectors from RGB images and associated 2D boxes. In specific, we propose three general components: prior injection module to obtain general object geometric priors from LLM model, 2D space projection constraint to minimize the discrepancy between the boundaries of projected 3D boxes and their corresponding 2D boxes on the image plane, and 3D space geometry constraint to build a Point-to-Box alignment loss to further refine the pose of estimated 3D boxes. Experiments on KITTI and SUN-RGBD datasets demonstrate that our method yields surprisingly high-quality 3D bounding boxes with only 2D annotation. The source code is available at https://github.com/gwenzhang/GGA., Comment: Accepted to ECCV24

Camera-based 3D occupancy prediction has recently garnered increasing attention in outdoor driving scenes. However, research in indoor scenes remains relatively unexplored. The core differences in indoor scenes lie in the complexity of scene scale and the variance in object size. In this paper, we propose a novel method, named ISO, for predicting indoor scene occupancy using monocular images. ISO harnesses the advantages of a pretrained depth model to achieve accurate depth predictions. Furthermore, we introduce the Dual Feature Line of Sight Projection (D-FLoSP) module within ISO, which enhances the learning of 3D voxel features. To foster further research in this domain, we introduce Occ-ScanNet, a large-scale occupancy benchmark for indoor scenes. With a dataset size 40 times larger than the NYUv2 dataset, it facilitates future scalable research in indoor scene analysis. Experimental results on both NYUv2 and Occ-ScanNet demonstrate that our method achieves state-of-the-art performance. The dataset and code are made publicly at https://github.com/hongxiaoy/ISO.git., Comment: Accepted by ECCV 2024

Serialization-based methods, which serialize the 3D voxels and group them into multiple sequences before inputting to Transformers, have demonstrated their effectiveness in 3D object detection. However, serializing 3D voxels into 1D sequences will inevitably sacrifice the voxel spatial proximity. Such an issue is hard to be addressed by enlarging the group size with existing serialization-based methods due to the quadratic complexity of Transformers with feature sizes. Inspired by the recent advances of state space models (SSMs), we present a Voxel SSM, termed as Voxel Mamba, which employs a group-free strategy to serialize the whole space of voxels into a single sequence. The linear complexity of SSMs encourages our group-free design, alleviating the loss of spatial proximity of voxels. To further enhance the spatial proximity, we propose a Dual-scale SSM Block to establish a hierarchical structure, enabling a larger receptive field in the 1D serialization curve, as well as more complete local regions in 3D space. Moreover, we implicitly apply window partition under the group-free framework by positional encoding, which further enhances spatial proximity by encoding voxel positional information. Our experiments on Waymo Open Dataset and nuScenes dataset show that Voxel Mamba not only achieves higher accuracy than state-of-the-art methods, but also demonstrates significant advantages in computational efficiency., Comment: 10 pages, 4 figures

End-to-end autonomous driving has garnered widespread attention. Current end-to-end approaches largely rely on the supervision from perception tasks such as detection, tracking, and map segmentation to aid in learning scene representations. However, these methods require extensive annotations, hindering the data scalability. To address this challenge, we propose a novel self-supervised method to enhance end-to-end driving without the need for costly labels. Specifically, our framework \textbf{LAW} uses a LAtent World model to predict future latent features based on the predicted ego actions and the latent feature of the current frame. The predicted latent features are supervised by the actually observed features in the future. This supervision jointly optimizes the latent feature learning and action prediction, which greatly enhances the driving performance. As a result, our approach achieves state-of-the-art performance in both open-loop and closed-loop benchmarks without costly annotations.

In this paper, we introduce Trim 3D Gaussian Splatting (TrimGS) to reconstruct accurate 3D geometry from images. Previous arts for geometry reconstruction from 3D Gaussians mainly focus on exploring strong geometry regularization. Instead, from a fresh perspective, we propose to obtain accurate 3D geometry of a scene by Gaussian trimming, which selectively removes the inaccurate geometry while preserving accurate structures. To achieve this, we analyze the contributions of individual 3D Gaussians and propose a contribution-based trimming strategy to remove the redundant or inaccurate Gaussians. Furthermore, our experimental and theoretical analyses reveal that a relatively small Gaussian scale is a non-negligible factor in representing and optimizing the intricate details. Therefore the proposed TrimGS maintains relatively small Gaussian scales. In addition, TrimGS is also compatible with the effective geometry regularization strategies in previous arts. When combined with the original 3DGS and the state-of-the-art 2DGS, TrimGS consistently yields more accurate geometry and higher perceptual quality. Our project page is https://trimgs.github.io, Comment: Project page: https://trimgs.github.io/

Single image reflection removal is inherently ambiguous, as both the reflection and transmission components requiring separation may follow natural image statistics. Existing methods attempt to address the issue by using various types of low-level and physics-based cues as sources of reflection signals. However, these cues are not universally applicable, since they are only observable in specific capture scenarios. This leads to a significant performance drop when test images do not align with their assumptions. In this paper, we aim to explore a novel flexible interactive reflection removal approach that leverages various forms of sparse human guidance, such as points and bounding boxes, as auxiliary high-level prior to achieve robust reflection removal. However, incorporating the raw user guidance naively into the existing reflection removal network does not result in performance gains. To this end, we innovatively transform raw user input into a unified form -- reflection masks using an Interactive Segmentation Foundation Model. Such a design absorbs the quintessence of the foundational segmentation model and flexible human guidance, thereby mitigating the challenges of reflection separations. Furthermore, to fully utilize user guidance and reduce user annotation costs, we design a mask-guided reflection removal network, comprising our proposed self-adaptive prompt block. This block adaptively incorporates user guidance as anchors and refines transmission features via cross-attention mechanisms. Extensive results on real-world images validate that our method demonstrates state-of-the-art performance on various datasets with the help of flexible and sparse user guidance. Our code and dataset will be publicly available here https://github.com/ShawnChenn/FlexibleReflectionRemoval.

Object-Centric Learning (OCL) seeks to enable Neural Networks to identify individual objects in visual scenes, which is crucial for interpretable visual comprehension and reasoning. Most existing OCL models adopt auto-encoding structures and learn to decompose visual scenes through specially designed inductive bias, which causes the model to miss small objects during reconstruction. Reverse hierarchy theory proposes that human vision corrects perception errors through a top-down visual pathway that returns to bottom-level neurons and acquires more detailed information, inspired by which we propose Reverse Hierarchy Guided Network (RHGNet) that introduces a top-down pathway that works in different ways in the training and inference processes. This pathway allows for guiding bottom-level features with top-level object representations during training, as well as encompassing information from bottom-level features into perception during inference. Our model achieves SOTA performance on several commonly used datasets including CLEVR, CLEVRTex and MOVi-C. We demonstrate with experiments that our method promotes the discovery of small objects and also generalizes well on complex real-world scenes. Code will be available at https://anonymous.4open.science/r/RHGNet-6CEF.

Thanks to the powerful generative capacity of diffusion models, recent years have witnessed rapid progress in human motion generation. Existing diffusion-based methods employ disparate network architectures and training strategies. The effect of the design of each component is still unclear. In addition, the iterative denoising process consumes considerable computational overhead, which is prohibitive for real-time scenarios such as virtual characters and humanoid robots. For this reason, we first conduct a comprehensive investigation into network architectures, training strategies, and inference processs. Based on the profound analysis, we tailor each component for efficient high-quality human motion generation. Despite the promising performance, the tailored model still suffers from foot skating which is an ubiquitous issue in diffusion-based solutions. To eliminate footskate, we identify foot-ground contact and correct foot motions along the denoising process. By organically combining these well-designed components together, we present StableMoFusion, a robust and efficient framework for human motion generation. Extensive experimental results show that our StableMoFusion performs favorably against current state-of-the-art methods. Project page: https://h-y1heng.github.io/StableMoFusion-page/

Recent advancements in generative models have significantly impacted content creation, leading to the emergence of Personalized Content Synthesis (PCS). With a small set of user-provided examples, PCS aims to customize the subject of interest to specific user-defined prompts. Over the past two years, more than 150 methods have been proposed. However, existing surveys mainly focus on text-to-image generation, with few providing up-to-date summaries on PCS. This paper offers a comprehensive survey of PCS, with a particular focus on the diffusion models. Specifically, we introduce the generic frameworks of PCS research, which can be broadly classified into optimization-based and learning-based approaches. We further categorize and analyze these methodologies, discussing their strengths, limitations, and key techniques. Additionally, we delve into specialized tasks within the field, such as personalized object generation, face synthesis, and style personalization, highlighting their unique challenges and innovations. Despite encouraging progress, we also present an analysis of the challenges such as overfitting and the trade-off between subject fidelity and text alignment. Through this detailed overview and analysis, we propose future directions to advance the development of PCS.

Image Anomaly Detection has been a challenging task in Computer Vision field. The advent of Vision-Language models, particularly the rise of CLIP-based frameworks, has opened new avenues for zero-shot anomaly detection. Recent studies have explored the use of CLIP by aligning images with normal and prompt descriptions. However, the exclusive dependence on textual guidance often falls short, highlighting the critical importance of additional visual references. In this work, we introduce a Dual-Image Enhanced CLIP approach, leveraging a joint vision-language scoring system. Our methods process pairs of images, utilizing each as a visual reference for the other, thereby enriching the inference process with visual context. This dual-image strategy markedly enhanced both anomaly classification and localization performances. Furthermore, we have strengthened our model with a test-time adaptation module that incorporates synthesized anomalies to refine localization capabilities. Our approach significantly exploits the potential of vision-language joint anomaly detection and demonstrates comparable performance with current SOTA methods across various datasets.

General world models represent a crucial pathway toward achieving Artificial General Intelligence (AGI), serving as the cornerstone for various applications ranging from virtual environments to decision-making systems. Recently, the emergence of the Sora model has attained significant attention due to its remarkable simulation capabilities, which exhibits an incipient comprehension of physical laws. In this survey, we embark on a comprehensive exploration of the latest advancements in world models. Our analysis navigates through the forefront of generative methodologies in video generation, where world models stand as pivotal constructs facilitating the synthesis of highly realistic visual content. Additionally, we scrutinize the burgeoning field of autonomous-driving world models, meticulously delineating their indispensable role in reshaping transportation and urban mobility. Furthermore, we delve into the intricacies inherent in world models deployed within autonomous agents, shedding light on their profound significance in enabling intelligent interactions within dynamic environmental contexts. At last, we examine challenges and limitations of world models, and discuss their potential future directions. We hope this survey can serve as a foundational reference for the research community and inspire continued innovation. This survey will be regularly updated at: https://github.com/GigaAI-research/General-World-Models-Survey., Comment: This survey will be regularly updated at: https://github.com/GigaAI-research/General-World-Models-Survey

Driven by powerful image diffusion models, recent research has achieved the automatic creation of 3D objects from textual or visual guidance. By performing score distillation sampling (SDS) iteratively across different views, these methods succeed in lifting 2D generative prior to the 3D space. However, such a 2D generative image prior bakes the effect of illumination and shadow into the texture. As a result, material maps optimized by SDS inevitably involve spurious correlated components. The absence of precise material definition makes it infeasible to relight the generated assets reasonably in novel scenes, which limits their application in downstream scenarios. In contrast, humans can effortlessly circumvent this ambiguity by deducing the material of the object from its appearance and semantics. Motivated by this insight, we propose MaterialSeg3D, a 3D asset material generation framework to infer underlying material from the 2D semantic prior. Based on such a prior model, we devise a mechanism to parse material in 3D space. We maintain a UV stack, each map of which is unprojected from a specific viewpoint. After traversing all viewpoints, we fuse the stack through a weighted voting scheme and then employ region unification to ensure the coherence of the object parts. To fuel the learning of semantics prior, we collect a material dataset, named Materialized Individual Objects (MIO), which features abundant images, diverse categories, and accurate annotations. Extensive quantitative and qualitative experiments demonstrate the effectiveness of our method.

Existing depth sensors are imperfect and may provide inaccurate depth values in challenging scenarios, such as in the presence of transparent or reflective objects. In this work, we present a general framework that leverages polarization imaging to improve inaccurate depth measurements from various depth sensors. Previous polarization-based depth enhancement methods focus on utilizing pure physics-based formulas for a single sensor. In contrast, our method first adopts a learning-based strategy where a neural network is trained to estimate a dense and complete depth map from polarization data and a sensor depth map from different sensors. To further improve the performance, we propose a Polarization Prompt Fusion Tuning (PPFT) strategy to effectively utilize RGB-based models pre-trained on large-scale datasets, as the size of the polarization dataset is limited to train a strong model from scratch. We conducted extensive experiments on a public dataset, and the results demonstrate that the proposed method performs favorably compared to existing depth enhancement baselines. Code and demos are available at https://lastbasket.github.io/PPFT/., Comment: CVPR 2024. Project page: https://lastbasket.github.io/PPFT/. The first two authors contribute equally

The advancement of real-time 3D scene reconstruction and novel view synthesis has been significantly propelled by 3D Gaussian Splatting (3DGS). However, effectively training large-scale 3DGS and rendering it in real-time across various scales remains challenging. This paper introduces CityGaussian (CityGS), which employs a novel divide-and-conquer training approach and Level-of-Detail (LoD) strategy for efficient large-scale 3DGS training and rendering. Specifically, the global scene prior and adaptive training data selection enables efficient training and seamless fusion. Based on fused Gaussian primitives, we generate different detail levels through compression, and realize fast rendering across various scales through the proposed block-wise detail levels selection and aggregation strategy. Extensive experimental results on large-scale scenes demonstrate that our approach attains state-of-theart rendering quality, enabling consistent real-time rendering of largescale scenes across vastly different scales. Our project page is available at https://dekuliutesla.github.io/citygs/., Comment: Accepted by ECCV2024; Project Page: https://dekuliutesla.github.io/citygs/

The remarkable performance of recent stereo depth estimation models benefits from the successful use of convolutional neural networks to regress dense disparity. Akin to most tasks, this needs gathering training data that covers a number of heterogeneous scenes at deployment time. However, training samples are typically acquired continuously in practical applications, making the capability to learn new scenes continually even more crucial. For this purpose, we propose to perform continual stereo matching where a model is tasked to 1) continually learn new scenes, 2) overcome forgetting previously learned scenes, and 3) continuously predict disparities at inference. We achieve this goal by introducing a Reusable Architecture Growth (RAG) framework. RAG leverages task-specific neural unit search and architecture growth to learn new scenes continually in both supervised and self-supervised manners. It can maintain high reusability during growth by reusing previous units while obtaining good performance. Additionally, we present a Scene Router module to adaptively select the scene-specific architecture path at inference. Comprehensive experiments on numerous datasets show that our framework performs impressively in various weather, road, and city circumstances and surpasses the state-of-the-art methods in more challenging cross-dataset settings. Further experiments also demonstrate the adaptability of our method to unseen scenes, which can facilitate end-to-end stereo architecture learning and practical deployment., Comment: Extended version of CVPR 2022 paper "Continual Stereo Matching of Continuous Driving Scenes with Growing Architecture" - Accepted to TPAMI in 2024

Continual learning (CL) aims to empower models to learn new tasks without forgetting previously acquired knowledge. Most prior works concentrate on the techniques of architectures, replay data, regularization, \etc. However, the category name of each class is largely neglected. Existing methods commonly utilize the one-hot labels and randomly initialize the classifier head. We argue that the scarce semantic information conveyed by the one-hot labels hampers the effective knowledge transfer across tasks. In this paper, we revisit the role of the classifier head within the CL paradigm and replace the classifier with semantic knowledge from pretrained language models (PLMs). Specifically, we use PLMs to generate semantic targets for each class, which are frozen and serve as supervision signals during training. Such targets fully consider the semantic correlation between all classes across tasks. Empirical studies show that our approach mitigates forgetting by alleviating representation drifting and facilitating knowledge transfer across tasks. The proposed method is simple to implement and can seamlessly be plugged into existing methods with negligible adjustments. Extensive experiments based on eleven mainstream baselines demonstrate the effectiveness and generalizability of our approach to various protocols. For example, under the class-incremental learning setting on ImageNet-100, our method significantly improves the Top-1 accuracy by 3.2\% to 6.1\% while reducing the forgetting rate by 2.6\% to 13.1\%., Comment: Accepted by CVPR 2024

Due to its great application potential, large-scale scene generation has drawn extensive attention in academia and industry. Recent research employs powerful generative models to create desired scenes and achieves promising results. However, most of these methods represent the scene using 3D primitives (e.g. point cloud or radiance field) incompatible with the industrial pipeline, which leads to a substantial gap between academic research and industrial deployment. Procedural Controllable Generation (PCG) is an efficient technique for creating scalable and high-quality assets, but it is unfriendly for ordinary users as it demands profound domain expertise. To address these issues, we resort to using the large language model (LLM) to drive the procedural modeling. In this paper, we introduce a large-scale scene generation framework, SceneX, which can automatically produce high-quality procedural models according to designers' textual descriptions.Specifically, the proposed method comprises two components, PCGBench and PCGPlanner. The former encompasses an extensive collection of accessible procedural assets and thousands of hand-craft API documents. The latter aims to generate executable actions for Blender to produce controllable and precise 3D assets guided by the user's instructions. Our SceneX can generate a city spanning 2.5 km times 2.5 km with delicate layout and geometric structures, drastically reducing the time cost from several weeks for professional PCG engineers to just a few hours for an ordinary user. Extensive experiments demonstrated the capability of our method in controllable large-scale scene generation and editing, including asset placement and season translation.

This paper presents a pilot study that explores the application of active learning, traditionally studied in the context of discriminative models, to generative models. We specifically focus on image synthesis personalization tasks. The primary challenge in conducting active learning on generative models lies in the open-ended nature of querying, which differs from the closed form of querying in discriminative models that typically target a single concept. We introduce the concept of anchor directions to transform the querying process into a semi-open problem. We propose a direction-based uncertainty sampling strategy to enable generative active learning and tackle the exploitation-exploration dilemma. Extensive experiments are conducted to validate the effectiveness of our approach, demonstrating that an open-source model can achieve superior performance compared to closed-source models developed by large companies, such as Google's StyleDrop. The source code is available at https://github.com/zhangxulu1996/GAL4Personalization.

For privacy and security concerns, the need to erase unwanted information from pre-trained vision models is becoming evident nowadays. In real-world scenarios, erasure requests originate at any time from both users and model owners. These requests usually form a sequence. Therefore, under such a setting, selective information is expected to be continuously removed from a pre-trained model while maintaining the rest. We define this problem as continual forgetting and identify two key challenges. (i) For unwanted knowledge, efficient and effective deleting is crucial. (ii) For remaining knowledge, the impact brought by the forgetting procedure should be minimal. To address them, we propose Group Sparse LoRA (GS-LoRA). Specifically, towards (i), we use LoRA modules to fine-tune the FFN layers in Transformer blocks for each forgetting task independently, and towards (ii), a simple group sparse regularization is adopted, enabling automatic selection of specific LoRA groups and zeroing out the others. GS-LoRA is effective, parameter-efficient, data-efficient, and easy to implement. We conduct extensive experiments on face recognition, object detection and image classification and demonstrate that GS-LoRA manages to forget specific classes with minimal impact on other classes. Codes will be released on \url{https://github.com/bjzhb666/GS-LoRA}., Comment: Accepted by CVPR 2024, latest version ahead of carema ready version

Machine learning has achieved remarkable success in many applications. However, existing studies are largely based on the closed-world assumption, which assumes that the environment is stationary, and the model is fixed once deployed. In many real-world applications, this fundamental and rather naive assumption may not hold because an open environment is complex, dynamic, and full of unknowns. In such cases, rejecting unknowns, discovering novelties, and then incrementally learning them, could enable models to be safe and evolve continually as biological systems do. This paper provides a holistic view of open-world machine learning by investigating unknown rejection, novel class discovery, and class-incremental learning in a unified paradigm. The challenges, principles, and limitations of current methodologies are discussed in detail. Finally, we discuss several potential directions for future research. This paper aims to provide a comprehensive introduction to the emerging open-world machine learning paradigm, to help researchers build more powerful AI systems in their respective fields, and to promote the development of artificial general intelligence.

Image-goal navigation is a challenging task that requires an agent to navigate to a goal indicated by an image in unfamiliar environments. Existing methods utilizing diverse scene memories suffer from inefficient exploration since they use all historical observations for decision-making without considering the goal-relevant fraction. To address this limitation, we present MemoNav, a novel memory model for image-goal navigation, which utilizes a working memory-inspired pipeline to improve navigation performance. Specifically, we employ three types of navigation memory. The node features on a map are stored in the short-term memory (STM), as these features are dynamically updated. A forgetting module then retains the informative STM fraction to increase efficiency. We also introduce long-term memory (LTM) to learn global scene representations by progressively aggregating STM features. Subsequently, a graph attention module encodes the retained STM and the LTM to generate working memory (WM) which contains the scene features essential for efficient navigation. The synergy among these three memory types boosts navigation performance by enabling the agent to learn and leverage goal-relevant scene features within a topological map. Our evaluation on multi-goal tasks demonstrates that MemoNav significantly outperforms previous methods across all difficulty levels in both Gibson and Matterport3D scenes. Qualitative results further illustrate that MemoNav plans more efficient routes., Comment: Accepted to CVPR 2024. Code: https://github.com/ZJULiHongxin/MemoNav

Speech-driven 3D facial animation is important for many multimedia applications. Recent work has shown promise in using either Diffusion models or Transformer architectures for this task. However, their mere aggregation does not lead to improved performance. We suspect this is due to a shortage of paired audio-4D data, which is crucial for the Transformer to effectively perform as a denoiser within the Diffusion framework. To tackle this issue, we present DiffSpeaker, a Transformer-based network equipped with novel biased conditional attention modules. These modules serve as substitutes for the traditional self/cross-attention in standard Transformers, incorporating thoughtfully designed biases that steer the attention mechanisms to concentrate on both the relevant task-specific and diffusion-related conditions. We also explore the trade-off between accurate lip synchronization and non-verbal facial expressions within the Diffusion paradigm. Experiments show our model not only achieves state-of-the-art performance on existing benchmarks, but also fast inference speed owing to its ability to generate facial motions in parallel., Comment: 9 pages, 5 figures. Code is avalable at https://github.com/theEricMa/DiffSpeaker

3D Gaussian Splatting has emerged as an alternative 3D representation for novel view synthesis, benefiting from its high-quality rendering results and real-time rendering speed. However, the 3D Gaussians learned by 3D-GS have ambiguous structures without any geometry constraints. This inherent issue in 3D-GS leads to a rough boundary when segmenting individual objects. To remedy these problems, we propose SAGD, a conceptually simple yet effective boundary-enhanced segmentation pipeline for 3D-GS to improve segmentation accuracy while preserving segmentation speed. Specifically, we introduce a Gaussian Decomposition scheme, which ingeniously utilizes the special structure of 3D Gaussian, finds out, and then decomposes the boundary Gaussians. Moreover, to achieve fast interactive 3D segmentation, we introduce a novel training-free pipeline by lifting a 2D foundation model to 3D-GS. Extensive experiments demonstrate that our approach achieves high-quality 3D segmentation without rough boundary issues, which can be easily applied to other scene editing tasks.

Label-efficient LiDAR-based 3D object detection is currently dominated by weakly/semi-supervised methods. Instead of exclusively following one of them, we propose MixSup, a more practical paradigm simultaneously utilizing massive cheap coarse labels and a limited number of accurate labels for Mixed-grained Supervision. We start by observing that point clouds are usually textureless, making it hard to learn semantics. However, point clouds are geometrically rich and scale-invariant to the distances from sensors, making it relatively easy to learn the geometry of objects, such as poses and shapes. Thus, MixSup leverages massive coarse cluster-level labels to learn semantics and a few expensive box-level labels to learn accurate poses and shapes. We redesign the label assignment in mainstream detectors, which allows them seamlessly integrated into MixSup, enabling practicality and universality. We validate its effectiveness in nuScenes, Waymo Open Dataset, and KITTI, employing various detectors. MixSup achieves up to 97.31% of fully supervised performance, using cheap cluster annotations and only 10% box annotations. Furthermore, we propose PointSAM based on the Segment Anything Model for automated coarse labeling, further reducing the annotation burden. The code is available at https://github.com/BraveGroup/PointSAM-for-MixSup., Comment: ICLR 2024, code is available at https://github.com/BraveGroup/PointSAM-for-MixSup

The text-to-image synthesis by diffusion models has recently shown remarkable performance in generating high-quality images. Although performs well for simple texts, the models may get confused when faced with complex texts that contain multiple objects or spatial relationships. To get the desired images, a feasible way is to manually adjust the textual descriptions, i.e., narrating the texts or adding some words, which is labor-consuming. In this paper, we propose a framework to learn the proper textual descriptions for diffusion models through prompt learning. By utilizing the quality guidance and the semantic guidance derived from the pre-trained diffusion model, our method can effectively learn the prompts to improve the matches between the input text and the generated images. Extensive experiments and analyses have validated the effectiveness of the proposed method.

Indoor scene generation has attracted significant attention recently as it is crucial for applications of gaming, virtual reality, and interior design. Current indoor scene generation methods can produce reasonable room layouts but often lack diversity and realism. This is primarily due to the limited coverage of existing datasets, including only large furniture without tiny furnishings in daily life. To address these challenges, we propose FurniScene, a large-scale 3D room dataset with intricate furnishing scenes from interior design professionals. Specifically, the FurniScene consists of 11,698 rooms and 39,691 unique furniture CAD models with 89 different types, covering things from large beds to small teacups on the coffee table. To better suit fine-grained indoor scene layout generation, we introduce a novel Two-Stage Diffusion Scene Model (TSDSM) and conduct an evaluation benchmark for various indoor scene generation based on FurniScene. Quantitative and qualitative evaluations demonstrate the capability of our method to generate highly realistic indoor scenes. Our dataset and code will be publicly available soon.

Recommender systems with cascading architecture play an increasingly significant role in online recommendation platforms, where the approach to dealing with negative feedback is a vital issue. For instance, in short video platforms, users tend to quickly slip away from candidates that they feel aversive, and recommender systems are expected to receive these explicit negative feedbacks and make adjustments to avoid these recommendations. Considering recency effect in memories, we propose a forgetting model based on Ebbinghaus Forgetting Curve to cope with negative feedback. In addition, we introduce a Pareto optimization solver to guarantee a better trade-off between recency and model performance. In conclusion, we propose Pareto-based Multi-Objective Recommender System with forgetting curve (PMORS), which can be applied to any multi-objective recommendation and show sufficiently superiority when facing explicit negative feedback. We have conducted evaluations of PMORS and achieved favorable outcomes in short-video scenarios on both public dataset and industrial dataset. After being deployed on an online short video platform named WeChat Channels in May, 2023, PMORS has not only demonstrated promising results for both consistency and recency but also achieved an improvement of up to +1.45% GMV.

Masked visual modeling has attracted much attention due to its promising potential in learning generalizable representations. Typical approaches urge models to predict specific contents of masked tokens, which can be intuitively considered as teaching a student (the model) to solve given problems (predicting masked contents). Under such settings, the performance is highly correlated with mask strategies (the difficulty of provided problems). We argue that it is equally important for the model to stand in the shoes of a teacher to produce challenging problems by itself. Intuitively, patches with high values of reconstruction loss can be regarded as hard samples, and masking those hard patches naturally becomes a demanding reconstruction task. To empower the model as a teacher, we propose Hard Patches Mining (HPM), predicting patch-wise losses and subsequently determining where to mask. Technically, we introduce an auxiliary loss predictor, which is trained with a relative objective to prevent overfitting to exact loss values. Also, to gradually guide the training procedure, we propose an easy-to-hard mask strategy. Empirically, HPM brings significant improvements under both image and video benchmarks. Interestingly, solely incorporating the extra loss prediction objective leads to better representations, verifying the efficacy of determining where is hard to reconstruct. The code is available at https://github.com/Haochen-Wang409/HPM., Comment: arXiv admin note: substantial text overlap with arXiv:2304.05919

Inversion methods, such as Textual Inversion, generate personalized images by incorporating concepts of interest provided by user images. However, existing methods often suffer from overfitting issues, where the dominant presence of inverted concepts leads to the absence of other desired concepts. It stems from the fact that during inversion, the irrelevant semantics in the user images are also encoded, forcing the inverted concepts to occupy locations far from the core distribution in the embedding space. To address this issue, we propose a method that guides the inversion process towards the core distribution for compositional embeddings. Additionally, we introduce a spatial regularization approach to balance the attention on the concepts being composed. Our method is designed as a post-training approach and can be seamlessly integrated with other inversion methods. Experimental results demonstrate the effectiveness of our proposed approach in mitigating the overfitting problem and generating more diverse and balanced compositions of concepts in the synthesized images. The source code is available at https://github.com/zhangxulu1996/Compositional-Inversion., Comment: This paper was accepted by AAAI 2024

Training Scene Graph Generation (SGG) models with natural language captions has become increasingly popular due to the abundant, cost-effective, and open-world generalization supervision signals that natural language offers. However, such unstructured caption data and its processing pose significant challenges in learning accurate and comprehensive scene graphs. The challenges can be summarized as three aspects: 1) traditional scene graph parsers based on linguistic representation often fail to extract meaningful relationship triplets from caption data. 2) grounding unlocalized objects of parsed triplets will meet ambiguity issues in visual-language alignment. 3) caption data typically are sparse and exhibit bias to partial observations of image content. Aiming to address these problems, we propose a divide-and-conquer strategy with a novel framework named \textit{GPT4SGG}, to obtain more accurate and comprehensive scene graph signals. This framework decomposes a complex scene into a bunch of simple regions, resulting in a set of region-specific narratives. With these region-specific narratives (partial observations) and a holistic narrative (global observation) for an image, a large language model (LLM) performs the relationship reasoning to synthesize an accurate and comprehensive scene graph. Experimental results demonstrate \textit{GPT4SGG} significantly improves the performance of SGG models trained on image-caption data, in which the ambiguity issue and long-tail bias have been well-handled with more accurate and comprehensive scene graphs.