Junran Peng

Yunjie Tian, Lingxi Xie, Jiemin Fang et al.

The past year has witnessed a rapid development of masked image modeling (MIM). MIM is mostly built upon the vision transformers, which suggests that self-supervised visual representations can be done by masking input image parts while requiring the target model to recover the missing contents. MIM has demonstrated promising results on downstream tasks, yet we are interested in whether there exist other effective ways to `learn by recovering missing contents'. In this paper, we investigate this topic by designing five other learning objectives that follow the same procedure as MIM but degrade the input image in different ways. With extensive experiments, we manage to summarize a few design principles for token-based pre-training of vision transformers. In particular, the best practice is obtained by keeping the original image style and enriching spatial masking with spatial misalignment -- this design achieves superior performance over MIM in a series of downstream recognition tasks without extra computational cost. The code is available at https://github.com/sunsmarterjie/beyond_masking.

Zekun Moore Wang, Zhongyuan Peng, Haoran Que et al.

The advent of Large Language Models (LLMs) has paved the way for complex tasks such as role-playing, which enhances user interactions by enabling models to imitate various characters. However, the closed-source nature of state-of-the-art LLMs and their general-purpose training limit role-playing optimization. In this paper, we introduce RoleLLM, a framework to benchmark, elicit, and enhance role-playing abilities in LLMs. RoleLLM comprises four stages: (1) Role Profile Construction for 100 roles; (2) Context-Based Instruction Generation (Context-Instruct) for role-specific knowledge extraction; (3) Role Prompting using GPT (RoleGPT) for speaking style imitation; and (4) Role-Conditioned Instruction Tuning (RoCIT) for fine-tuning open-source models along with role customization. By Context-Instruct and RoleGPT, we create RoleBench, the first systematic and fine-grained character-level benchmark dataset for role-playing with 168,093 samples. Moreover, RoCIT on RoleBench yields RoleLLaMA (English) and RoleGLM (Chinese), significantly enhancing role-playing abilities and even achieving comparable results with RoleGPT (using GPT-4).

Qing Chang, Junran Peng, Lingxie Xie et al.

The paradigm of training models on massive data without label through self-supervised learning (SSL) and finetuning on many downstream tasks has become a trend recently. However, due to the high training costs and the unconsciousness of downstream usages, most self-supervised learning methods lack the capability to correspond to the diversities of downstream scenarios, as there are various data domains, different vision tasks and latency constraints on models. Neural architecture search (NAS) is one universally acknowledged fashion to conquer the issues above, but applying NAS on SSL seems impossible as there is no label or metric provided for judging model selection. In this paper, we present DATA, a simple yet effective NAS approach specialized for SSL that provides Domain-Aware and Task-Aware pre-training. Specifically, we (i) train a supernet which could be deemed as a set of millions of networks covering a wide range of model scales without any label, (ii) propose a flexible searching mechanism compatible with SSL that enables finding networks of different computation costs, for various downstream vision tasks and data domains without explicit metric provided. Instantiated With MoCo v2, our method achieves promising results across a wide range of computation costs on downstream tasks, including image classification, object detection and semantic segmentation. DATA is orthogonal to most existing SSL methods and endows them the ability of customization on downstream needs. Extensive experiments on other SSL methods demonstrate the generalizability of the proposed method. Code is released at https://github.com/GAIA-vision/GAIA-ssl

Haoran Que, Feiyu Duan, Liqun He et al.

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.

Hongcheng Guo, Jian Yang, Jiaheng Liu et al.

With the rapid development of IT operations, it has become increasingly crucial to efficiently manage and analyze large volumes of data for practical applications. The techniques of Natural Language Processing (NLP) have shown remarkable capabilities for various tasks, including named entity recognition, machine translation and dialogue systems. Recently, Large Language Models (LLMs) have achieved significant improvements across various NLP downstream tasks. However, there is a lack of specialized LLMs for IT operations. In this paper, we introduce the OWL, a large language model trained on our collected OWL-Instruct dataset with a wide range of IT-related information, where the mixture-of-adapter strategy is proposed to improve the parameter-efficient tuning across different domains or tasks. Furthermore, we evaluate the performance of our OWL on the OWL-Bench established by us and open IT-related benchmarks. OWL demonstrates superior performance results on IT tasks, which outperforms existing models by significant margins. Moreover, we hope that the findings of our work will provide more insights to revolutionize the techniques of IT operations with specialized LLMs.

Lingfeng Li, Huaiwei Cong, Gangming Zhao et al.

In recent years, several works have adopted the convolutional neural network (CNN) to diagnose the avascular necrosis of the femoral head (AVNFH) based on X-ray images or magnetic resonance imaging (MRI). However, due to the tissue overlap, X-ray images are difficult to provide fine-grained features for early diagnosis. MRI, on the other hand, has a long imaging time, is more expensive, making it impractical in mass screening. Computed tomography (CT) shows layer-wise tissues, is faster to image, and is less costly than MRI. However, to our knowledge, there is no work on CT-based automated diagnosis of AVNFH. In this work, we collected and labeled a large-scale dataset for AVNFH ranking. In addition, existing end-to-end CNNs only yields the classification result and are difficult to provide more information for doctors in diagnosis. To address this issue, we propose the structure regularized attentive network (SRANet), which is able to highlight the necrotic regions during classification based on patch attention. SRANet extracts features in chunks of images, obtains weight via the attention mechanism to aggregate the features, and constrains them by a structural regularizer with prior knowledge to improve the generalization. SRANet was evaluated on our AVNFH-CT dataset. Experimental results show that SRANet is superior to CNNs for AVNFH classification, moreover, it can localize lesions and provide more information to assist doctors in diagnosis. Our codes are made public at https://github.com/tomas-lilingfeng/SRANet.

Shougao Zhang, Mengqi Zhou, Yuxi Wang et al.

Urban areas, as the primary human habitat in modern civilization, accommodate a broad spectrum of social activities. With the surge of embodied intelligence, recent years have witnessed an increasing presence of physical agents in urban areas, such as autonomous vehicles and delivery robots. As a result, practitioners significantly value crafting authentic, simulation-ready 3D cities to facilitate the training and verification of such agents. However, this task is quite challenging. Current generative methods fall short in either diversity, controllability, or fidelity. In this work, we resort to the procedural content generation (PCG) technique for high-fidelity generation. It assembles superior assets according to empirical rules, ultimately leading to industrial-grade outcomes. To ensure diverse and self contained creation, we design a management protocol to accommodate extensive PCG plugins with distinct functions and interfaces. Based on this unified PCG library, we develop a multi-agent framework to transform multi-modal instructions, including OSM, semantic maps, and satellite images, into executable programs. The programs coordinate relevant plugins to construct the 3D city consistent with the control condition. A visual feedback scheme is introduced to further refine the initial outcomes. Our method, named CityX, demonstrates its superiority in creating diverse, controllable, and realistic 3D urban scenes. The synthetic scenes can be seamlessly deployed as a real-time simulator and an infinite data generator for embodied intelligence research. Our project page: https://cityx-lab.github.io.

Liqun Yang, Jian Yang, Chaoren Wei et al.

Fuzzing is an important dynamic program analysis technique designed for finding vulnerabilities in complex software. Fuzzing involves presenting a target program with crafted malicious input to cause crashes, buffer overflows, memory errors, and exceptions. Crafting malicious inputs in an efficient manner is a difficult open problem and the best approaches often apply uniform random mutations to pre-existing valid inputs. In this work, we propose to adopt fine-tuned large language models (FuzzCoder) to learn patterns in the input files from successful attacks to guide future fuzzing explorations. Specifically, we develop a framework to leverage the code LLMs to guide the mutation process of inputs in fuzzing. The mutation process is formulated as the sequence-to-sequence modeling, where LLM receives a sequence of bytes and then outputs the mutated byte sequence. FuzzCoder is fine-tuned on the created instruction dataset (Fuzz-Instruct), where the successful fuzzing history is collected from the heuristic fuzzing tool. FuzzCoder can predict mutation locations and strategies locations in input files to trigger abnormal behaviors of the program. Experimental results show that FuzzCoder based on AFL (American Fuzzy Lop) gain significant improvements in terms of effective proportion of mutation (EPM) and number of crashes (NC) for various input formats including ELF, JPG, MP3, and XML.

Xin Wei, Huaiwei Cong, Zheng Zhang et al.

Long-term vertebral fractures severely affect the life quality of patients, causing kyphotic, lumbar deformity and even paralysis. Computed tomography (CT) is a common clinical examination to screen for this disease at early stages. However, the faint radiological appearances and unspecific symptoms lead to a high risk of missed diagnosis, especially for the mild vertebral fractures. In this paper, we argue that reinforcing the faint fracture features to encourage the inter-class separability is the key to improving the accuracy. Motivated by this, we propose a supervised contrastive learning based model to estimate Genent's Grade of vertebral fracture with CT scans. The supervised contrastive learning, as an auxiliary task, narrows the distance of features within the same class while pushing others away, enhancing the model's capability of capturing subtle features of vertebral fractures. Our method has a specificity of 99% and a sensitivity of 85% in binary classification, and a macro-F1 of 77% in multi-class classification, indicating that contrastive learning significantly improves the accuracy of vertebrae fracture screening. Considering the lack of datasets in this field, we construct a database including 208 samples annotated by experienced radiologists. Our desensitized data and codes will be made publicly available for the community.

Yuxue Yang, Lue Fan, Ziqi Shi et al.

In this paper, we propose NeoVerse, a versatile 4D world model that is capable of 4D reconstruction, novel-trajectory video generation, and rich downstream applications. We first identify a common limitation of scalability in current 4D world modeling methods, caused either by expensive and specialized multi-view 4D data or by cumbersome training pre-processing. In contrast, our NeoVerse is built upon a core philosophy that makes the full pipeline scalable to diverse in-the-wild monocular videos. Specifically, NeoVerse features pose-free feed-forward 4D reconstruction, online monocular degradation pattern simulation, and other well-aligned techniques. These designs empower NeoVerse with versatility and generalization to various domains. Meanwhile, NeoVerse achieves state-of-the-art performance in standard reconstruction and generation benchmarks. Our project page is available at https://neoverse-4d.github.io

Pei Wang, Yanan Wu, Zekun Wang et al.

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.

Junran Peng, Yiheng Huang, Silei Shen et al.

In this paper, we introduce RoleMotion, a large-scale human motion dataset that encompasses a wealth of role-playing and functional motion data tailored to fit various specific scenes. Existing text datasets are mainly constructed decentrally as amalgamation of assorted subsets that their data are nonfunctional and isolated to work together to cover social activities in various scenes. Also, the quality of motion data is inconsistent, and textual annotation lacks fine-grained details in these datasets. In contrast, RoleMotion is meticulously designed and collected with a particular focus on scenes and roles. The dataset features 25 classic scenes, 110 functional roles, over 500 behaviors, and 10296 high-quality human motion sequences of body and hands, annotated with 27831 fine-grained text descriptions. We build an evaluator stronger than existing counterparts, prove its reliability, and evaluate various text-to-motion methods on our dataset. Finally, we explore the interplay of motion generation of body and hands. Experimental results demonstrate the high-quality and functionality of our dataset on text-driven whole-body generation.

Caorui Li, Yu Chen, Yiyan Ji et al. · pku

Recent advances in multimodal large language models (MLLMs) have demonstrated substantial potential in video understanding. However, existing benchmarks fail to comprehensively evaluate synergistic reasoning capabilities across audio and visual modalities, often neglecting either one of the modalities or integrating them in a logically inconsistent manner. To bridge this gap, we introduce OmniVideoBench, a large-scale and rigorously designed benchmark dedicated to assessing synergistic audio-visual understanding, with a strong emphasis on modality complementarity and logical consistency. Specifically, OmniVideoBench comprises 1000 high-quality question-answer(QA) pairs, each annotated with step-by-step reasoning traces, derived from 628 diverse videos ranging from several seconds to 30 minutes, and manually verified to guarantee complete correctness and uniqueness. Moreover, OmniVideoBench encompasses 13 carefully designed question types, covering temporal reasoning, spatial localization, counting, causal inference, summarization, and beyond, thereby capturing the essential challenges of video understanding. Evaluation of multiple MLLMs on OmniVideoBench reveals a pronounced gap between model performance and human reasoning, with open-source models lagging significantly behind their closed-source counterparts, underscoring the inherent difficulty of genuine audio-visual reasoning. We will release OmniVideoBench to foster the development of MLLMs with stronger and more generalizable reasoning capabilities.

Yang Liu, He Guan, Chuanchen Luo et al.

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/.

Zehao Ni, Yonghao He, Lingfeng Qian et al.

In the context of imitation learning, visuomotor-based diffusion policy learning is one of the main directions in robotic manipulation. Most of these approaches rely on point clouds as observation inputs and construct scene representations through point clouds feature learning, which enables them to achieve remarkable accuracy. However, the existing literature lacks an in-depth exploration of vision-only solutions that have significant potential. In this paper, we propose a Vision-Only and single-view Diffusion Policy learning method (VO-DP) that leverages pretrained visual foundation models to achieve effective fusion of semantic and geometric features. We utilize intermediate features from VGGT incorporating semantic features from DINOv2 and geometric features from Alternating Attention blocks. Features are fused via cross-attention and spatially compressed with a CNN to form the input to the policy head. Extensive experiments demonstrate that VO-DP not only outperforms the vision-only baseline DP significantly but also exhibits distinct performance trends against the point cloud-based method DP3: in simulation tasks, VO-DP achieves an average success rate of 64.6% on par with DP3 64.0% and far higher than DP 34.8%, while in real-world tasks, it reaches 87.9%, outperforming both DP3 67.5% and DP 11.2% by a notable margin. Further robustness evaluations confirm that VO-DP remains highly stable under varying conditions including color, size, background, and lighting. Lastly, we open-source a training library for robotic manipulation. Built on Accelerate, this library supports multi-machine and multi-GPU parallel training, as well as mixed precision training. It is compatible with visuomotor policies such as DP, DP3 and VO-DP, and also supports the RoboTwin simulator.

Hongcheng Guo, Jian Yang, Jiaheng Liu et al.

Log anomaly detection is a key component in the field of artificial intelligence for IT operations (AIOps). Considering log data of variant domains, retraining the whole network for unknown domains is inefficient in real industrial scenarios. However, previous deep models merely focused on extracting the semantics of log sequences in the same domain, leading to poor generalization on multi-domain logs. To alleviate this issue, we propose a unified Transformer-based framework for Log anomaly detection (LogFormer) to improve the generalization ability across different domains, where we establish a two-stage process including the pre-training and adapter-based tuning stage. Specifically, our model is first pre-trained on the source domain to obtain shared semantic knowledge of log data. Then, we transfer such knowledge to the target domain via shared parameters. Besides, the Log-Attention module is proposed to supplement the information ignored by the log-paring. The proposed method is evaluated on three public and one real-world datasets. Experimental results on multiple benchmarks demonstrate the effectiveness of our LogFormer with fewer trainable parameters and lower training costs.

Yang Liu, Chuanchen Luo, Zhongkai Mao et al.

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/.

Xu Hu, Yuxi Wang, Lue Fan et al.

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.

Genghao Zhang, Yuxi Wang, Chuanchen Luo et al.

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.

Chang Yu, Junran Peng, Xiangyu Zhu et al.

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.

Yixuan Zhang, Hui Yang, Chuanchen Luo et al.

Generating realistic 3D human-object interactions (HOIs) from text descriptions is a active research topic with potential applications in virtual and augmented reality, robotics, and animation. However, creating high-quality 3D HOIs remains challenging due to the lack of large-scale interaction data and the difficulty of ensuring physical plausibility, especially in out-of-domain (OOD) scenarios. Current methods tend to focus either on the body or the hands, which limits their ability to produce cohesive and realistic interactions. In this paper, we propose OOD-HOI, a text-driven framework for generating whole-body human-object interactions that generalize well to new objects and actions. Our approach integrates a dual-branch reciprocal diffusion model to synthesize initial interaction poses, a contact-guided interaction refiner to improve physical accuracy based on predicted contact areas, and a dynamic adaptation mechanism which includes semantic adjustment and geometry deformation to improve robustness. Experimental results demonstrate that our OOD-HOI could generate more realistic and physically plausible 3D interaction pose in OOD scenarios compared to existing methods.

Xu Shi, Wei Yao, Chuanchen Luo et al.

Recently, significant progress has been made in text-based motion generation, enabling the generation of diverse and high-quality human motions that conform to textual descriptions. However, generating motions beyond the distribution of original datasets remains challenging, i.e., zero-shot generation. By adopting a divide-and-conquer strategy, we propose a new framework named Fine-Grained Human Motion Diffusion Model (FG-MDM) for zero-shot human motion generation. Specifically, we first parse previous vague textual annotations into fine-grained descriptions of different body parts by leveraging a large language model. We then use these fine-grained descriptions to guide a transformer-based diffusion model, which further adopts a design of part tokens. FG-MDM can generate human motions beyond the scope of original datasets owing to descriptions that are closer to motion essence. Our experimental results demonstrate the superiority of FG-MDM over previous methods in zero-shot settings. We will release our fine-grained textual annotations for HumanML3D and KIT.

Zeyu Li, Ruitong Gan, Chuanchen Luo et al.

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.

Yang Liu, Chuanchen Luo, Zimo Tang et al.

Illumination and texture editing are critical dimensions for world-to-world transfer, which is valuable for applications including sim2real and real2real visual data scaling up for embodied AI. Existing techniques generatively re-render the input video to realize the transfer, such as video relighting models and conditioned world generation models. Nevertheless, these models are predominantly limited to the domain of training data (e.g., portrait) or fall into the bottleneck of temporal consistency and computation efficiency, especially when the input video involves complex dynamics and long durations. In this paper, we propose TC-Light, a novel generative renderer to overcome these problems. Starting from the video preliminarily relighted by an inflated video relighting model, it optimizes appearance embedding in the first stage to align global illumination. Then it optimizes the proposed canonical video representation, i.e., Unique Video Tensor (UVT), to align fine-grained texture and lighting in the second stage. To comprehensively evaluate performance, we also establish a long and highly dynamic video benchmark. Extensive experiments show that our method enables physically plausible re-rendering results with superior temporal coherence and low computation cost. The code and video demos are available at https://dekuliutesla.github.io/tclight/.

Ruitong Gan, Junran Peng, Yang Liu et al.

Planes are fundamental primitives of 3D sences, especially in man-made environments such as indoor spaces and urban streets. Representing these planes in a structured and parameterized format facilitates scene editing and physical simulations in downstream applications. Recently, Gaussian Splatting (GS) has demonstrated remarkable effectiveness in the Novel View Synthesis task, with extensions showing great potential in accurate surface reconstruction. However, even state-of-the-art GS representations often struggle to reconstruct planar regions with sufficient smoothness and precision. To address this issue, we propose GSPlane, which recovers accurate geometry and produces clean and well-structured mesh connectivity for plane regions in the reconstructed scene. By leveraging off-the-shelf segmentation and normal prediction models, GSPlane extracts robust planar priors to establish structured representations for planar Gaussian coordinates, which help guide the training process by enforcing geometric consistency. To further enhance training robustness, a Dynamic Gaussian Re-classifier is introduced to adaptively reclassify planar Gaussians with persistently high gradients as non-planar, ensuring more reliable optimization. Furthermore, we utilize the optimized planar priors to refine the mesh layouts, significantly improving topological structure while reducing the number of vertices and faces. We also explore applications of the structured planar representation, which enable decoupling and flexible manipulation of objects on supportive planes. Extensive experiments demonstrate that, with no sacrifice in rendering quality, the introduction of planar priors significantly improves the geometric accuracy of the extracted meshes across various baselines.

Yang Liu, Chuanchen Luo, Zimo Tang et al.

We study the problem of applying 3D Foundation Models (3DFMs) to dense Novel View Synthesis (NVS). Despite significant progress in Novel View Synthesis powered by NeRF and 3DGS, current approaches remain reliant on accurate 3D attributes (e.g., camera poses and point clouds) acquired from Structure-from-Motion (SfM), which is often slow and fragile in low-texture or low-overlap captures. Recent 3DFMs showcase orders of magnitude speedup over the traditional pipeline and great potential for online NVS. But most of the validation and conclusions are confined to sparse-view settings. Our study reveals that naively scaling 3DFMs to dense views encounters two fundamental barriers: dramatically increasing VRAM burden and imperfect outputs that degrade initialization-sensitive 3D training. To address these barriers, we introduce VGGT-X, incorporating a memory-efficient VGGT implementation that scales to 1,000+ images, an adaptive global alignment for VGGT output enhancement, and robust 3DGS training practices. Extensive experiments show that these measures substantially close the fidelity gap with COLMAP-initialized pipelines, achieving state-of-the-art results in dense COLMAP-free NVS and pose estimation. Additionally, we analyze the causes of remaining gaps with COLMAP-initialized rendering, providing insights for the future development of 3D foundation models and dense NVS. Our project page is available at https://dekuliutesla.github.io/vggt-x.github.io/

Yixuan Zhang, Qing Chang, Yuxi Wang et al.

Speech-driven 3D facial animation seeks to produce lifelike facial expressions that are synchronized with the speech content and its emotional nuances, finding applications in various multimedia fields. However, previous methods often overlook emotional facial expressions or fail to disentangle them effectively from the speech content. To address these challenges, we present EmoDiffusion, a novel approach that disentangles different emotions in speech to generate rich 3D emotional facial expressions. Specifically, our method employs two Variational Autoencoders (VAEs) to separately generate the upper face region and mouth region, thereby learning a more refined representation of the facial sequence. Unlike traditional methods that use diffusion models to connect facial expression sequences with audio inputs, we perform the diffusion process in the latent space. Furthermore, we introduce an Emotion Adapter to evaluate upper face movements accurately. Given the paucity of 3D emotional talking face data in the animation industry, we capture facial expressions under the guidance of animation experts using LiveLinkFace on an iPhone. This effort results in the creation of an innovative 3D blendshape emotional talking face dataset (3D-BEF) used to train our network. Extensive experiments and perceptual evaluations validate the effectiveness of our approach, confirming its superiority in generating realistic and emotionally rich facial animations.

M-A-P Team, Xinrun Du, Yifan Yao et al.

Large language models (LLMs) have demonstrated remarkable proficiency in mainstream academic disciplines such as mathematics, physics, and computer science. However, human knowledge encompasses over 200 specialized disciplines, far exceeding the scope of existing benchmarks. The capabilities of LLMs in many of these specialized fields-particularly in light industry, agriculture, and service-oriented disciplines-remain inadequately evaluated. To address this gap, we present SuperGPQA, a comprehensive benchmark that evaluates graduate-level knowledge and reasoning capabilities across 285 disciplines. Our benchmark employs a novel Human-LLM collaborative filtering mechanism to eliminate trivial or ambiguous questions through iterative refinement based on both LLM responses and expert feedback. Our experimental results reveal significant room for improvement in the performance of current state-of-the-art LLMs across diverse knowledge domains (e.g., the reasoning-focused model DeepSeek-R1 achieved the highest accuracy of 61.82% on SuperGPQA), highlighting the considerable gap between current model capabilities and artificial general intelligence. Additionally, we present comprehensive insights from our management of a large-scale annotation process, involving over 80 expert annotators and an interactive Human-LLM collaborative system, offering valuable methodological guidance for future research initiatives of comparable scope.

Yiheng Huang, Hui Yang, Chuanchen Luo et al.

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/

Mengqi Zhou, Yuxi Wang, Jun Hou et al.

Developing comprehensive explicit world models is crucial for understanding and simulating real-world scenarios. Recently, Procedural Controllable Generation (PCG) has gained significant attention in large-scale scene generation by enabling the creation of scalable, high-quality assets. However, PCG faces challenges such as limited modular diversity, high expertise requirements, and challenges in managing the diverse elements and structures in complex scenes. 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, PCGHub and PCGPlanner. The former encompasses an extensive collection of accessible procedural assets and thousands of hand-craft API documents to perform as a standard protocol for PCG controller. The latter aims to generate executable actions for Blender to produce controllable and precise 3D assets guided by the user's instructions. Extensive experiments demonstrated the capability of our method in controllable large-scale scene generation, including nature scenes and unbounded cities, as well as scene editing such as asset placement and season translation.

Xingyuan Bu, Junran Peng, Junjie Yan et al.

Transfer learning with pre-training on large-scale datasets has played an increasingly significant role in computer vision and natural language processing recently. However, as there exist numerous application scenarios that have distinctive demands such as certain latency constraints and specialized data distributions, it is prohibitively expensive to take advantage of large-scale pre-training for per-task requirements. In this paper, we focus on the area of object detection and present a transfer learning system named GAIA, which could automatically and efficiently give birth to customized solutions according to heterogeneous downstream needs. GAIA is capable of providing powerful pre-trained weights, selecting models that conform to downstream demands such as latency constraints and specified data domains, and collecting relevant data for practitioners who have very few datapoints for their tasks. With GAIA, we achieve promising results on COCO, Objects365, Open Images, Caltech, CityPersons, and UODB which is a collection of datasets including KITTI, VOC, WiderFace, DOTA, Clipart, Comic, and more. Taking COCO as an example, GAIA is able to efficiently produce models covering a wide range of latency from 16ms to 53ms, and yields AP from 38.2 to 46.5 without whistles and bells. To benefit every practitioner in the community of object detection, GAIA is released at https://github.com/GAIA-vision.

Junran Peng, Xingyuan Bu, Ming Sun et al.

Training with more data has always been the most stable and effective way of improving performance in deep learning era. As the largest object detection dataset so far, Open Images brings great opportunities and challenges for object detection in general and sophisticated scenarios. However, owing to its semi-automatic collecting and labeling pipeline to deal with the huge data scale, Open Images dataset suffers from label-related problems that objects may explicitly or implicitly have multiple labels and the label distribution is extremely imbalanced. In this work, we quantitatively analyze these label problems and provide a simple but effective solution. We design a concurrent softmax to handle the multi-label problems in object detection and propose a soft-sampling methods with hybrid training scheduler to deal with the label imbalance. Overall, our method yields a dramatic improvement of 3.34 points, leading to the best single model with 60.90 mAP on the public object detection test set of Open Images. And our ensembling result achieves 67.17 mAP, which is 4.29 points higher than the best result of Open Images public test 2018.

Xingyuan Bu, Junran Peng, Changbao Wang et al.

This report details our solution to the Google AI Open Images Challenge 2019 Object Detection Track. Based on our detailed analysis on the Open Images dataset, it is found that there are four typical features: large-scale, hierarchical tag system, severe annotation incompleteness and data imbalance. Considering these characteristics, many strategies are employed, including larger backbone, distributed softmax loss, class-aware sampling, expert model, and heavier classifier. In virtue of these effective strategies, our best single model could achieve a mAP of 61.90. After ensemble, the final mAP is boosted to 67.17 in the public leaderboard and 64.21 in the private leaderboard, which earns 3rd place in the Open Images Challenge 2019.

Junran Peng, Ming Sun, Zhaoxiang Zhang et al.

Recently, Neural Architecture Search has achieved great success in large-scale image classification. In contrast, there have been limited works focusing on architecture search for object detection, mainly because the costly ImageNet pre-training is always required for detectors. Training from scratch, as a substitute, demands more epochs to converge and brings no computation saving. To overcome this obstacle, we introduce a practical neural architecture transformation search(NATS)algorithm for object detection in this paper. Instead of searching and constructing an entire network, NATS explores the architecture space on the base of existing network and reusing its weights. We propose a novel neural architecture search strategy in channel-level instead of path-level and devise a search space specially targeting at object detection. With the combination of these two designs, an architecture transformation scheme could be discovered to adapt a network designed for image classification to task of object detection. Since our method is gradient-based and only searches for a transformation scheme, the weights of models pretrained inImageNet could be utilized in both searching and retraining stage, which makes the whole process very efficient. The transformed network requires no extra parameters and FLOPs, and is friendly to hardware optimization, which is practical to use in real-time application. In experiments, we demonstrate the effectiveness of NATSon networks like ResNet and ResNeXt. Our transformed networks, combined with various detection frameworks, achieve significant improvements on the COCO dataset while keeping fast.

Junran Peng, Ming Sun, Zhaoxiang Zhang et al.

Scale-sensitive object detection remains a challenging task, where most of the existing methods could not learn it explicitly and are not robust to scale variance. In addition, the most existing methods are less efficient during training or slow during inference, which are not friendly to real-time applications. In this paper, we propose a practical object detection method with scale-sensitive network.Our method first predicts a global continuous scale ,which is shared by all position, for each convolution filter of each network stage. To effectively learn the scale, we average the spatial features and distill the scale from channels. For fast-deployment, we propose a scale decomposition method that transfers the robust fractional scale into combination of fixed integral scales for each convolution filter, which exploits the dilated convolution. We demonstrate it on one-stage and two-stage algorithms under different configurations. For practical applications, training of our method is of efficiency and simplicity which gets rid of complex data sampling or optimize strategy. During test-ing, the proposed method requires no extra operation and is very supportive of hardware acceleration like TensorRT and TVM. On the COCO test-dev, our model could achieve a 41.5 mAP on one-stage detector and 42.1 mAP on two-stage detectors based on ResNet-101, outperforming base-lines by 2.4 and 2.1 respectively without extra FLOPS.

Junran Peng, Lingxi Xie, Zhaoxiang Zhang et al.

This paper presents an efficient module named spatial bottleneck for accelerating the convolutional layers in deep neural networks. The core idea is to decompose convolution into two stages, which first reduce the spatial resolution of the feature map, and then restore it to the desired size. This operation decreases the sampling density in the spatial domain, which is independent yet complementary to network acceleration approaches in the channel domain. Using different sampling rates, we can tradeoff between recognition accuracy and model complexity. As a basic building block, spatial bottleneck can be used to replace any single convolutional layer, or the combination of two convolutional layers. We empirically verify the effectiveness of spatial bottleneck by applying it to the deep residual networks. Spatial bottleneck achieves 2x and 1.4x speedup on the regular and channel-bottlenecked residual blocks, respectively, with the accuracies retained in recognizing low-resolution images, and even improved in recognizing high-resolution images.