Zhihui Li

Jinjie Shen, Yaxiong Wang, Yujiao Wu et al.

The rapid rise of generative AI has made multimodal fake news increasingly realistic and pervasive, posing severe threats to public trust and social stability. Existing detection methods rely heavily on manipulation-specific models and large-scale labeled data, resulting in poor generalization to emerging manipulation types. We observed that the essence of manipulated misinformation lies in its intrinsic conflicts, \textbf{i.e.,} semantic or physical inconsistencies either across modalities or with common world knowledge. Inspired by this observation, we propose \textbf{C}onflict-\textbf{O}riented \textbf{RE}asoning (\textbf{CORE}) framework, an effective paradigm that learns to endows multimodal large language models (MLLMs) with explicit conflict-capturing capability. To this end, CORE first constructs the Conflict Attribution Corpus (CAC) with fine-grained annotations of conflict factors and sources, providing essential data support for subsequent conflict perception training. By performing conflict-oriented representation enhancement and reasoning based on CAC, CORE achieves robust and generalizable conflict detection, effectively and rapidly adapting to unseen manipulation types with a few samples or in even zero-shot settings. Extensive experiments demonstrate that CORE surpasses state-of-the-art models. The dataset and code are publicly available at https://github.com/shen8424/CORE.

Siyi Hu, Yifan Zhong, Minquan Gao et al.

A significant challenge facing researchers in the area of multi-agent reinforcement learning (MARL) pertains to the identification of a library that can offer fast and compatible development for multi-agent tasks and algorithm combinations, while obviating the need to consider compatibility issues. In this paper, we present MARLlib, a library designed to address the aforementioned challenge by leveraging three key mechanisms: 1) a standardized multi-agent environment wrapper, 2) an agent-level algorithm implementation, and 3) a flexible policy mapping strategy. By utilizing these mechanisms, MARLlib can effectively disentangle the intertwined nature of the multi-agent task and the learning process of the algorithm, with the ability to automatically alter the training strategy based on the current task's attributes. The MARLlib library's source code is publicly accessible on GitHub: \url{https://github.com/Replicable-MARL/MARLlib}.

Longtao Jiang, Jiangmin Bao, Zhendong Wang et al.

Normalizing flows (NFs) provide exact likelihoods and deterministic invertible sampling, but have historically lagged behind diffusion models for large-scale image generation. We identify a key obstacle: NFs are required to learn a single invertible transport over the full ambient space, making them highly sensitive to high-dimensional representations. This leads to a semantic-capacity mismatch in modern visual representation spaces, where semantic information is compact but encoded in overcomplete features. We propose SRC-Flow, which introduces a Semantic Representation Compressor (SRC) to compact high-dimensional RAE features into a low-dimensional semantic space before flow modeling and preserve reconstruction through the frozen RAE decoder. This compact space reduces the modeling burden of NFs and enables effective likelihood-based generation in semantic representation space. We further adopt constant noise regularization tailored to the fixed unconditional bijection learned by flows. On ImageNet $256 \times 256$ and $512 \times 512$, SRC-Flow achieves state-of-the-art generation quality among normalizing flow methods, with gFID scores of 1.65 and 2.07 under classifier-free guidance, while retaining exact likelihood computation in the compact semantic representation space and deterministic invertible sampling at the flow level. Codes and models will be available at https://github.com/longtaojiang/SRC-Flow.

Zizhuo Lin, Quanling Liu, Jinsheng Quan et al.

Large language models (LLMs) often solve a task when all instructions are given in a single prompt, but fail when the same information is revealed gradually across turns. When a clean FULL prompt and a RAW-SHARDED conversation contain the same complete user evidence, the model should still arrive at the same answer. We argue that a key reason for this gap is self-anchored drift: responses produced under partial information introduce unsupported assumptions, and those assumptions later distort the final answer. To reduce this effect, we propose Canonical-Context On-Policy Distillation (CCOPD). During training, the same base model is used in two roles: a frozen teacher conditioned on the clean FULL prompt and a trainable student that receives the same evidence incrementally through a multi-turn conversation; CCOPD aligns the student's behavior on its own trajectories with the teacher's canonical full-context behavior. Trained only on math problem conversations, CCOPD yields a 32\% average relative improvement in RAW-SHARDED performance over the original base model across math and five zero-shot out-of-domain task families, while largely preserving full-context performance. Further analyses suggest that CCOPD strengthens grounding in user evidence and reduces sensitivity to contamination from earlier assistant turns.

Kecheng Zhang, Zongxin Yang, Mingfei Han et al.

Visual agents operating in the wild must respond to queries precisely when sufficient evidence first appears in a video stream, a critical capability that is overlooked by conventional video LLMs evaluated in offline settings. The shift to an online, streaming paradigm introduces significant challenges: a lack of decision transparency, the difficulty of aligning response timing with visual evidence, and the need to maintain a global, causally consistent understanding under tight computational budgets. To address these issues, we propose a novel framework that decouples reasoning control from memory integration. We introduce \textbf{\model{}}, an instantiation of this framework with two core components. First, the \emph{Active Thinking Decision Maker (ATDM)} is a transparent reasoning controller that externalizes its decision process using observable progress ($\boldsymbolρ$) and confidence ($\boldsymbol{c}$) metrics. This allows it to precisely time its response $t_r$ to match the first-sufficient-evidence timestamp $t^\star$ while streaming its reasoning to the user. Second, the \emph{Hierarchical Progressive Semantic Integration (HPSI)} module acts as an efficient memory system. It employs a set of learnable, multi-level aggregation tokens that are propagated across clips to build a rich, global cognitive state without exceeding token budgets. %Our approach sets a new standard on key online video understanding benchmarks, achieving strong performance of \textbf{71.6\%} on StreamingBench and \textbf{46.9\%} on OVOBench, demonstrating a robust solution for evidence-aligned and transparent online video analysis. Extensive experiments demonstrate the effectiveness of ATDM and HPSI, e.g., Thinking-QwenVL improves the accuracy of the previous state-of-the-art from 67.63\% to 71.60\% on the StreamingBench benchmark.

Hu Gao, Zhihui Li, Depeng Dang et al.

Spacecraft pose estimation plays a vital role in many on-orbit space missions, such as rendezvous and docking, debris removal, and on-orbit maintenance. At present, space images contain widely varying lighting conditions, high contrast and low resolution, pose estimation of space objects is more challenging than that of objects on earth. In this paper, we analyzing the radar image characteristics of spacecraft on-orbit, then propose a new deep learning neural Network structure named Dense Residual U-shaped Network (DR-U-Net) to extract image features. We further introduce a novel neural network based on DR-U-Net, namely Spacecraft U-shaped Network (SU-Net) to achieve end-to-end pose estimation for non-cooperative spacecraft. Specifically, the SU-Net first preprocess the image of non-cooperative spacecraft, then transfer learning was used for pre-training. Subsequently, in order to solve the problem of radar image blur and low ability of spacecraft contour recognition, we add residual connection and dense connection to the backbone network U-Net, and we named it DR-U-Net. In this way, the feature loss and the complexity of the model is reduced, and the degradation of deep neural network during training is avoided. Finally, a layer of feedforward neural network is used for pose estimation of non-cooperative spacecraft on-orbit. Experiments prove that the proposed method does not rely on the hand-made object specific features, and the model has robust robustness, and the calculation accuracy outperforms the state-of-the-art pose estimation methods. The absolute error is 0.1557 to 0.4491 , the mean error is about 0.302 , and the standard deviation is about 0.065 .

Hu Gao, Zhihui Li, Depeng Dang et al.

Uncontrolled spacecraft will disintegrate and generate a large amount of debris in the reentry process, and ablative debris may cause potential risks to the safety of human life and property on the ground. Therefore, predicting the landing points of spacecraft debris and forecasting the degree of risk of debris to human life and property is very important. In view that it is difficult to predict the process of reentry process and the reentry point in advance, and the debris generated from reentry disintegration may cause ground damage for the uncontrolled space vehicle on expiration of service. In this paper, we adopt the object-oriented approach to consider the spacecraft and its disintegrated components as consisting of simple basic geometric models, and introduce three machine learning models: the support vector regression (SVR), decision tree regression (DTR) and multilayer perceptron (MLP) to predict the velocity, longitude and latitude of spacecraft debris landing points for the first time. Then, we compare the prediction accuracy of the three models. Furthermore, we define the reentry risk and the degree of danger, and we calculate the risk level for each spacecraft debris and make warnings accordingly. The experimental results show that the proposed method can obtain high accuracy prediction results in at least 15 seconds and make safety level warning more real-time.

Tingjun Dai, Mingfei Han, Tingwen Du et al.

Measurement of task progress through explicit, actionable milestones is critical for robust robotic manipulation. This progress awareness enables a model to ground its current task status, anticipate verifiable intermediate states, and detect and recover from failures when progress stalls. To embody this capability, we introduce See, Plan, Rewind (SPR), a progress-aware vision-language-action framework that dynamically grounds language instructions into a sequence of spatial subgoals. SPR operates through a continuous core cycle, Seeing the current state and upcoming milestone, Planning a trajectory towards the next 2D waypoint, and Rewinding to a recoverable state upon failure by monitoring progress against the expected sequence. This closed-loop approach enables robust error correction without requiring additional training data or auxiliary models. Extensive experiments demonstrate the framework's effectiveness, generalization and robustness: SPR outperforms the MolmoAct baseline by 5\% on the LIBERO benchmark. On the challenging LIBERO-Plus benchmark with unseen instructions and initial states, SPR achieves state-of-the-art robustness with the smallest performance drop, surpassing OpenVLA-OFT and UniVLA, demonstrating superior out-of-distribution robustness.

Zhihui Li, Francesco Montomoli, Sanjiv Sharma

In this study, we utilize the emerging Physics Informed Neural Networks (PINNs) approach for the first time to predict the flow field of a compressor cascade. Different from conventional training methods, a new adaptive learning strategy that mitigates gradient imbalance through incorporating adaptive weights in conjunction with dynamically adjusting learning rate is used during the training process to improve the convergence of PINNs. The performance of PINNs is assessed here by solving both the forward and inverse problems. In the forward problem, by encapsulating the physical relations among relevant variables, PINNs demonstrate their effectiveness in accurately forecasting the compressor's flow field. PINNs also show obvious advantages over the traditional CFD approaches, particularly in scenarios lacking complete boundary conditions, as is often the case in inverse engineering problems. PINNs successfully reconstruct the flow field of the compressor cascade solely based on partial velocity vectors and near-wall pressure information. Furthermore, PINNs show robust performance in the environment of various levels of aleatory uncertainties stemming from labeled data. This research provides evidence that PINNs can offer turbomachinery designers an additional and promising option alongside the current dominant CFD methods.

Changlin Li, Jiawei Zhang, Shuhao Liu et al.

Human video generation has advanced rapidly with the development of diffusion models, but the high computational cost and substantial memory consumption associated with training these models on high-resolution, multi-frame data pose significant challenges. In this paper, we propose Entropy-Guided Prioritized Progressive Learning (Ent-Prog), an efficient training framework tailored for diffusion models on human video generation. First, we introduce Conditional Entropy Inflation (CEI) to assess the importance of different model components on the target conditional generation task, enabling prioritized training of the most critical components. Second, we introduce an adaptive progressive schedule that adaptively increases computational complexity during training by measuring the convergence efficiency. Ent-Prog reduces both training time and GPU memory consumption while maintaining model performance. Extensive experiments across three datasets, demonstrate the effectiveness of Ent-Prog, achieving up to 2.2$\times$ training speedup and 2.4$\times$ GPU memory reduction without compromising generative performance.

Mind Lab, Song Cao, Vic Cao et al.

We present MindLab Toolkit (MinT), a managed infrastructure system for Low-Rank Adaptation (LoRA) post-training and online serving. MinT targets a setting where many trained policies are produced over a small number of expensive base-model deployments. Instead of materializing each policy as a merged full checkpoint, MinT keeps the base model resident and moves exported LoRA adapter revisions through rollout, update, export, evaluation, serving, and rollback, hiding distributed training, serving, scheduling, and data movement behind a service interface. MinT scales this path along three axes. Scale Up extends LoRA RL to frontier-scale dense and MoE architectures, including MLA and DSA attention paths, with training and serving validated beyond 1T total parameters. Scale Down moves only the exported LoRA adapter, which can be under 1% of base-model size in rank-1 settings; adapter-only handoff reduces the measured step by 18.3x on a 4B dense model and 2.85x on a 30B MoE, while concurrent multi-policy GRPO shortens wall time by 1.77x and 1.45x without raising peak memory. Scale Out separates durable policy addressability from CPU/GPU working sets: a tensor-parallel deployment supports 10^6-scale addressable catalogs (measured single-engine sweeps through 100K) and thousand-adapter active waves at cluster scale, with cold loading treated as scheduled service work and packed MoE LoRA tensors improving live engine loading by 8.5-8.7x. MinT thus manages million-scale LoRA policy catalogs while training and serving selected adapter revisions over shared 1T-class base models.

Haihong Hao, Lei Chen, Mingfei Han et al.

Existing vision-and-language navigation (VLN) models primarily reason over past and current visual observations, while largely ignoring the future visual dynamics induced by actions. As a result, they often lack an effective understanding of the causal relationship between actions and how the visual world changes, limiting robust decision-making. Humans, in contrast, can imagine the near future by leveraging action-dynamics causality, which improves both environmental understanding and navigation choices. Inspired by this capability, we propose LatentPilot, a new paradigm that exploits future observations during training as a valuable data source to learn action-conditioned visual dynamics, while requiring no access to future frames at inference. Concretely, we propose a flywheel-style training mechanism that iteratively collects on-policy trajectories and retrains the model to better match the agent's behavior distribution, with an expert takeover triggered when the agent deviates excessively. LatentPilot further learns visual latent tokens without explicit supervision; these latent tokens attend globally in a continuous latent space and are carried across steps, serving as both the current output and the next input, thereby enabling the agent to dream ahead and reason about how actions will affect subsequent observations. Experiments on R2R-CE, RxR-CE, and R2R-PE benchmarks achieve new SOTA results, and real-robot tests across diverse environments demonstrate LatentPilot's superior understanding of environment-action dynamics in scene. Project page:https://abdd.top/latentpilot/

Changlin Li, Jiawei Zhang, Zeyi Shi et al.

Large-scale vision generative models, including diffusion and flow models, have demonstrated remarkable performance in visual generation tasks. However, transferring these pre-trained models to downstream tasks often results in significant parameter redundancy. In this paper, we propose EntPruner, an entropy-guided automatic progressive pruning framework for diffusion and flow models. First, we introduce entropy-guided pruning, a block-level importance assessment strategy specifically designed for generative models. Unlike discriminative models, generative models require preserving the diversity and condition-fidelity of the output distribution. As the importance of each module can vary significantly across downstream tasks, EntPruner prioritizes pruning of less important blocks using data-dependent Conditional Entropy Deviation (CED) as a guiding metric. CED quantifies how much the distribution diverges from the learned conditional data distribution after removing a block. Second, we propose a zero-shot adaptive pruning framework to automatically determine when and how much to prune during training. This dynamic strategy avoids the pitfalls of one-shot pruning, mitigating mode collapse, and preserving model performance. Extensive experiments on DiT and SiT models demonstrate the effectiveness of EntPruner, achieving up to 2.22$\times$ inference speedup while maintaining competitive generation quality on ImageNet and three downstream datasets.

Longtao Jiang, Jie Huang, Mingfei Han et al.

Text-guided image inpainting aims to inpaint masked image regions based on a textual prompt while preserving the background. Although diffusion-based methods have become dominant, their property of modeling the entire image in latent space makes it challenging for the results to align well with prompt details and maintain a consistent background. To address these issues, we explore Mask AutoRegressive (MAR) models for this task. MAR naturally supports image inpainting by generating latent tokens corresponding to mask regions, enabling better local controllability without altering the background. However, directly applying MAR to this task makes the inpainting content either ignore the prompts or be disharmonious with the background context. Through analysis of the attention maps from the inpainting images, we identify the impact of background tokens on text tokens during the MAR generation, and leverage this to design \textbf{Token Painter}, a training-free text-guided image inpainting method based on MAR. Our approach introduces two key components: (1) Dual-Stream Encoder Information Fusion (DEIF), which fuses the semantic and context information from text and background in frequency domain to produce novel guidance tokens, allowing MAR to generate text-faithful inpainting content while keeping harmonious with background context. (2) Adaptive Decoder Attention Score Enhancing (ADAE), which adaptively enhances attention scores on guidance tokens and inpainting tokens to further enhance the alignment of prompt details and the content visual quality. Extensive experiments demonstrate that our training-free method outperforms prior state-of-the-art methods across almost all metrics. Codes: https://github.com/longtaojiang/Token-Painter.

Changlin Li, Guangrun Wang, Bing Wang et al.

Dynamic networks have shown their promising capability in reducing theoretical computation complexity by adapting their architectures to the input during inference. However, their practical runtime usually lags behind the theoretical acceleration due to inefficient sparsity. Here, we explore a hardware-efficient dynamic inference regime, named dynamic weight slicing, which adaptively slice a part of network parameters for inputs with diverse difficulty levels, while keeping parameters stored statically and contiguously in hardware to prevent the extra burden of sparse computation. Based on this scheme, we present dynamic slimmable network (DS-Net) and dynamic slice-able network (DS-Net++) by input-dependently adjusting filter numbers of CNNs and multiple dimensions in both CNNs and transformers, respectively. To ensure sub-network generality and routing fairness, we propose a disentangled two-stage optimization scheme with training techniques such as in-place bootstrapping (IB), multi-view consistency (MvCo) and sandwich gate sparsification (SGS) to train supernet and gate separately. Extensive experiments on 4 datasets and 3 different network architectures demonstrate our method consistently outperforms state-of-the-art static and dynamic model compression methods by a large margin (up to 6.6%). Typically, DS-Net++ achieves 2-4x computation reduction and 1.62x real-world acceleration over MobileNet, ResNet-50 and Vision Transformer, with minimal accuracy drops (0.1-0.3%) on ImageNet. Code release: https://github.com/changlin31/DS-Net

Changlin Li, Guangrun Wang, Bing Wang et al.

Current dynamic networks and dynamic pruning methods have shown their promising capability in reducing theoretical computation complexity. However, dynamic sparse patterns on convolutional filters fail to achieve actual acceleration in real-world implementation, due to the extra burden of indexing, weight-copying, or zero-masking. Here, we explore a dynamic network slimming regime, named Dynamic Slimmable Network (DS-Net), which aims to achieve good hardware-efficiency via dynamically adjusting filter numbers of networks at test time with respect to different inputs, while keeping filters stored statically and contiguously in hardware to prevent the extra burden. Our DS-Net is empowered with the ability of dynamic inference by the proposed double-headed dynamic gate that comprises an attention head and a slimming head to predictively adjust network width with negligible extra computation cost. To ensure generality of each candidate architecture and the fairness of gate, we propose a disentangled two-stage training scheme inspired by one-shot NAS. In the first stage, a novel training technique for weight-sharing networks named In-place Ensemble Bootstrapping is proposed to improve the supernet training efficacy. In the second stage, Sandwich Gate Sparsification is proposed to assist the gate training by identifying easy and hard samples in an online way. Extensive experiments demonstrate our DS-Net consistently outperforms its static counterparts as well as state-of-the-art static and dynamic model compression methods by a large margin (up to 5.9%). Typically, DS-Net achieves 2-4x computation reduction and 1.62x real-world acceleration over ResNet-50 and MobileNet with minimal accuracy drops on ImageNet. Code release: https://github.com/changlin31/DS-Net .

Yiyang Zou, Tianhao Zhao, Peilun Xiao et al.

Accident anticipation aims to predict impending collisions from dashcam videos and trigger early alerts. Existing methods rely on binary supervision with manually annotated "anomaly onset" frames, which are subjective and inconsistent, leading to inaccurate risk estimation. In contrast, we propose RiskProp, a novel collision-anchored self-supervised risk propagation paradigm for early accident anticipation, which removes the need for anomaly onset annotations and leverages only the reliably annotated collision frame. RiskProp models temporal risk evolution through two observation-driven losses: first, since future frames contain more definitive evidence of an impending accident, we introduce a future-frame regularization loss that uses the model's next-frame prediction as a soft target to supervise the current frame, enabling backward propagation of risk signals; second, inspired by the empirical trend of rising risk before accidents, we design an adaptive monotonic constraint to encourage a non-decreasing progression over time. Experiments on CAP and Nexar demonstrate that RiskProp achieves state-of-the-art performance and produces smoother, more discriminative risk curves, improving both early anticipation and interpretability.

Jinxing Zhou, Zhihui Li, Yongqiang Yu et al.

We present \textbf{Met}a-\textbf{T}oken \textbf{Le}arning (Mettle), a simple and memory-efficient method for adapting large-scale pretrained transformer models to downstream audio-visual tasks. Instead of sequentially modifying the output feature distribution of the transformer backbone, Mettle utilizes a lightweight \textit{Layer-Centric Distillation (LCD)} module to distill in parallel the intact audio or visual features embedded by each transformer layer into compact meta-tokens. This distillation process considers both pretrained knowledge preservation and task-specific adaptation. The obtained meta-tokens can be directly applied to classification tasks, such as audio-visual event localization and audio-visual video parsing. To further support fine-grained segmentation tasks, such as audio-visual segmentation, we introduce a \textit{Meta-Token Injection (MTI)} module, which utilizes the audio and visual meta-tokens distilled from the top transformer layer to guide feature adaptation in earlier layers. Extensive experiments on multiple audiovisual benchmarks demonstrate that our method significantly reduces memory usage and training time while maintaining parameter efficiency and competitive accuracy.

Minghao Jin, Mozheng Liao, Mingfei Han et al.

Recent world-model-based Vision-Language-Action (VLA) architectures have improved robotic manipulation through predictive visual foresight. However, dense future prediction introduces visual redundancy and accumulates errors, causing long-horizon plan drift. Meanwhile, recent sparse methods typically represent visual foresight using high-level semantic subtasks or implicit latent states. These representations often lack explicit kinematic grounding, weakening the alignment between planning and low-level execution. To address this, we propose StructVLA, which reformulates a generative world model into an explicit structured planner for reliable control. Instead of dense rollouts or semantic goals, StructVLA predicts sparse, physically meaningful structured frames. Derived from intrinsic kinematic cues (e.g., gripper transitions and kinematic turning points), these frames capture spatiotemporal milestones closely aligned with task progress. We implement this approach through a two-stage training paradigm with a unified discrete token vocabulary: the world model is first trained to predict structured frames and subsequently optimized to map the structured foresight into low-level actions. This approach provides clear physical guidance and bridges visual planning and motion control. In our experiments, StructVLA achieves strong average success rates of 75.0% on SimplerEnv-WidowX and 94.8% on LIBERO. Real-world deployments further demonstrate reliable task completion and robust generalization across both basic pick-and-place and complex long-horizon tasks.

Haihong Hao, Mingfei Han, Changlin Li et al.

Embodied navigation demands comprehensive scene understanding and precise spatial reasoning. While image-text models excel at interpreting pixel-level color and lighting cues, 3D-text models capture volumetric structure and spatial relationships. However, unified fusion approaches that jointly fuse 2D images, 3D point clouds, and textual instructions face challenges in limited availability of triple-modality data and difficulty resolving conflicting beliefs among modalities. In this work, we introduce CoNav, a collaborative cross-modal reasoning framework where a pretrained 3D-text model explicitly guides an image-text navigation agent by providing structured spatial-semantic knowledge to resolve ambiguities during navigation. Specifically, we introduce Cross-Modal Belief Alignment, which operationalizes this cross-modal guidance by simply sharing textual hypotheses from the 3D-text model to the navigation agent. Through lightweight fine-tuning on a small 2D-3D-text corpus, the navigation agent learns to integrate visual cues with spatial-semantic knowledge derived from the 3D-text model, enabling effective reasoning in embodied navigation. CoNav achieves significant improvements on four standard embodied navigation benchmarks (R2R, CVDN, REVERIE, SOON) and two spatial reasoning benchmarks (ScanQA, SQA3D). Moreover, under close navigation Success Rate, CoNav often generates shorter paths compared to other methods (as measured by SPL), showcasing the potential and challenges of fusing data from different modalities in embodied navigation. Project Page: https://oceanhao.github.io/CoNav/

Li Yu, Zhihui Li, Jimin Xiao et al.

Neural Representations for Videos (NeRV) have simplified the video codec process and achieved swift decoding speeds by encoding video content into a neural network, presenting a promising solution for video compression. However, existing work overlooks the crucial issue that videos reconstructed by these methods lack high-frequency details. To address this problem, this paper introduces a High-Frequency Enhanced Hybrid Neural Representation Network. Our method focuses on leveraging high-frequency information to improve the synthesis of fine details by the network. Specifically, we design a wavelet high-frequency encoder that incorporates Wavelet Frequency Decomposer (WFD) blocks to generate high-frequency feature embeddings. Next, we design the High-Frequency Feature Modulation (HFM) block, which leverages the extracted high-frequency embeddings to enhance the fitting process of the decoder. Finally, with the refined Harmonic decoder block and a Dynamic Weighted Frequency Loss, we further reduce the potential loss of high-frequency information. Experiments on the Bunny and UVG datasets demonstrate that our method outperforms other methods, showing notable improvements in detail preservation and compression performance.

Mingfei Han, Haihong Hao, Jinxing Zhou et al.

Vision-language models often hallucinate details, generating non-existent objects or inaccurate attributes that compromise output reliability. Existing methods typically address these issues via extensive human annotations or external supervision from more powerful models. In this work, we present a novel framework that leverages the model's self-consistency between long responses and short answers to generate preference pairs for training. We observe that short binary questions tend to yield highly reliable responses, which can be used to query the target model to evaluate and rank its generated responses. Specifically, we design a self-reflection pipeline where detailed model responses are compared against concise binary answers, and inconsistency signals are utilized to automatically curate high-quality training data without human annotations or external model-based supervision. By relying solely on self-consistency rather than external supervision, our method offers a scalable and efficient solution that effectively reduces hallucinations using unlabeled data. Extensive experiments on multiple benchmarks, i.e., AMBER, MultiObject-Hal (ROPE), Object HalBench, and MMHal-Bench, demonstrate significant improvements in factual grounding and reliability. Moreover, our approach maintains robust instruction-following ability, as evidenced by enhanced performance on LLaVA-Bench and MMBench.

Changliang Xia, Chengyou Jia, Zhuohang Dang et al.

Dense prediction tasks hold significant importance of computer vision, aiming to learn pixel-wise annotated labels for input images. Despite advances in this field, existing methods primarily focus on idealized conditions, exhibiting limited real-world generalization and struggling with the acute scarcity of real-world data in practical scenarios. To systematically study this problem, we first introduce DenseWorld, a benchmark spanning a broad set of 25 dense prediction tasks that correspond to urgent real-world applications, featuring unified evaluation across tasks. We then propose DenseDiT, which exploits generative models' visual priors to perform diverse real-world dense prediction tasks through a unified strategy. DenseDiT combines a parameter-reuse mechanism and two lightweight branches that adaptively integrate multi-scale context. This design enables DenseDiT to achieve efficient tuning with less than 0.1% additional parameters, activating the visual priors while effectively adapting to diverse real-world dense prediction tasks. Evaluations on DenseWorld reveal significant performance drops in existing general and specialized baselines, highlighting their limited real-world generalization. In contrast, DenseDiT achieves superior results using less than 0.01% training data of baselines, underscoring its practical value for real-world deployment.

Li Yu, Zhihui Li, Chao Yao et al.

Neural representations for video (NeRV) have gained considerable attention for their strong performance across various video tasks. However, existing NeRV methods often struggle to capture fine spatial details, resulting in vague reconstructions. In this paper, we present a Frequency Separation and Augmentation based Neural Representation for video (FANeRV), which addresses these limitations with its core Wavelet Frequency Upgrade Block. This block explicitly separates input frames into high and low-frequency components using discrete wavelet transform, followed by targeted enhancement using specialized modules. Finally, a specially designed gated network effectively fuses these frequency components for optimal reconstruction. Additionally, convolutional residual enhancement blocks are integrated into the later stages of the network to balance parameter distribution and improve the restoration of high-frequency details. Experimental results demonstrate that FANeRV significantly improves reconstruction performance and excels in multiple tasks, including video compression, inpainting, and interpolation, outperforming existing NeRV methods.

Sihao Lin, Daqi Liu, Ruochong Fu et al.

Estimating the 3D world from 2D monocular images is a fundamental yet challenging task due to the labour-intensive nature of 3D annotations. To simplify label acquisition, this work proposes a novel approach that bridges 2D vision foundation models (VFMs) with 3D tasks by decoupling 3D supervision into an ensemble of image-level primitives, e.g., semantic and geometric components. As a key motivator, we leverage the zero-shot capabilities of vision-language models for image semantics. However, due to the notorious ill-posed problem - multiple distinct 3D scenes can produce identical 2D projections, directly inferring metric depth from a monocular image in a zero-shot manner is unsuitable. In contrast, 2D VFMs provide promising sources of relative depth, which theoretically aligns with metric depth when properly scaled and offset. Thus, we adapt the relative depth derived from VFMs into metric depth by optimising the scale and offset using temporal consistency, also known as novel view synthesis, without access to ground-truth metric depth. Consequently, we project the semantics into 3D space using the reconstructed metric depth, thereby providing 3D supervision. Extensive experiments on nuScenes and SemanticKITTI demonstrate the effectiveness of our framework. For instance, the proposed method surpasses the current state-of-the-art by 3.34% mIoU on nuScenes for voxel occupancy prediction.

Meng Zhang, Zhihui Li, Zhibin Yu

In the context of global warming, even relatively cooler countries like the UK are experiencing a rise in cooling demand, particularly in southern regions such as London. This growing demand, especially during the summer months, presents significant challenges for energy management systems. Accurately predicting cooling demand in urban domestic buildings is essential for maintaining energy efficiency. This study introduces a generalised framework for developing high-resolution Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks using physical model-based summer cooling demand data. To maximise the predictive capability and generalisation ability of the models under limited data scenarios, four distinct data partitioning strategies were implemented, including the extrapolation, month-based interpolation, global interpolation, and day-based interpolation. Bayesian Optimisation (BO) was then applied to fine-tune the hyper-parameters, substantially improving the framework predictive accuracy. Results show that the day-based interpolation GRU model demonstrated the best performance due to its ability to retain both the data randomness and the time sequence continuity characteristics. This optimal model achieves a Root Mean Squared Error (RMSE) of 2.22%, a Mean Absolute Error (MAE) of 0.87%, and a coefficient of determination (R square) of 0.9386 on the test set. The generalisation ability of this framework was further evaluated by forecasting.

De Cheng, Xiaojian Huang, Nannan Wang et al.

Unsupervised learning visible-infrared person re-identification (USL-VI-ReID) aims at learning modality-invariant features from unlabeled cross-modality dataset, which is crucial for practical applications in video surveillance systems. The key to essentially address the USL-VI-ReID task is to solve the cross-modality data association problem for further heterogeneous joint learning. To address this issue, we propose a Dual Optimal Transport Label Assignment (DOTLA) framework to simultaneously assign the generated labels from one modality to its counterpart modality. The proposed DOTLA mechanism formulates a mutual reinforcement and efficient solution to cross-modality data association, which could effectively reduce the side-effects of some insufficient and noisy label associations. Besides, we further propose a cross-modality neighbor consistency guided label refinement and regularization module, to eliminate the negative effects brought by the inaccurate supervised signals, under the assumption that the prediction or label distribution of each example should be similar to its nearest neighbors. Extensive experimental results on the public SYSU-MM01 and RegDB datasets demonstrate the effectiveness of the proposed method, surpassing existing state-of-the-art approach by a large margin of 7.76% mAP on average, which even surpasses some supervised VI-ReID methods.

Xiaojun Chang, Pengzhen Ren, Pengfei Xu et al.

Scene graph is a structured representation of a scene that can clearly express the objects, attributes, and relationships between objects in the scene. As computer vision technology continues to develop, people are no longer satisfied with simply detecting and recognizing objects in images; instead, people look forward to a higher level of understanding and reasoning about visual scenes. For example, given an image, we want to not only detect and recognize objects in the image, but also know the relationship between objects (visual relationship detection), and generate a text description (image captioning) based on the image content. Alternatively, we might want the machine to tell us what the little girl in the image is doing (Visual Question Answering (VQA)), or even remove the dog from the image and find similar images (image editing and retrieval), etc. These tasks require a higher level of understanding and reasoning for image vision tasks. The scene graph is just such a powerful tool for scene understanding. Therefore, scene graphs have attracted the attention of a large number of researchers, and related research is often cross-modal, complex, and rapidly developing. However, no relatively systematic survey of scene graphs exists at present. To this end, this survey conducts a comprehensive investigation of the current scene graph research. More specifically, we first summarized the general definition of the scene graph, then conducted a comprehensive and systematic discussion on the generation method of the scene graph (SGG) and the SGG with the aid of prior knowledge. We then investigated the main applications of scene graphs and summarized the most commonly used datasets. Finally, we provide some insights into the future development of scene graphs. We believe this will be a very helpful foundation for future research on scene graphs.

Pengzhen Ren, Gang Xiao, Xiaojun Chang et al.

In the field of complex action recognition in videos, the quality of the designed model plays a crucial role in the final performance. However, artificially designed network structures often rely heavily on the researchers' knowledge and experience. Accordingly, because of the automated design of its network structure, Neural architecture search (NAS) has achieved great success in the image processing field and attracted substantial research attention in recent years. Although some NAS methods have reduced the number of GPU search days required to single digits in the image field, directly using 3D convolution to extend NAS to the video field is still likely to produce a surge in computing volume. To address this challenge, we propose a new processing framework called Neural Architecture Search- Temporal Convolutional (NAS-TC). Our proposed framework is divided into two phases. In the first phase, the classical CNN network is used as the backbone network to complete the computationally intensive feature extraction task. In the second stage, a simple stitching search to the cell is used to complete the relatively lightweight long-range temporal-dependent information extraction. This ensures our method will have more reasonable parameter assignments and can handle minute-level videos. Finally, we conduct sufficient experiments on multiple benchmark datasets and obtain competitive recognition accuracy.

Caixia Yan, Xiaojun Chang, Minnan Luo et al.

Linear discriminant analysis (LDA) is a popular technique to learn the most discriminative features for multi-class classification. A vast majority of existing LDA algorithms are prone to be dominated by the class with very large deviation from the others, i.e., edge class, which occurs frequently in multi-class classification. First, the existence of edge classes often makes the total mean biased in the calculation of between-class scatter matrix. Second, the exploitation of l2-norm based between-class distance criterion magnifies the extremely large distance corresponding to edge class. In this regard, a novel self-weighted robust LDA with l21-norm based pairwise between-class distance criterion, called SWRLDA, is proposed for multi-class classification especially with edge classes. SWRLDA can automatically avoid the optimal mean calculation and simultaneously learn adaptive weights for each class pair without setting any additional parameter. An efficient re-weighted algorithm is exploited to derive the global optimum of the challenging l21-norm maximization problem. The proposed SWRLDA is easy to implement, and converges fast in practice. Extensive experiments demonstrate that SWRLDA performs favorably against other compared methods on both synthetic and real-world datasets, while presenting superior computational efficiency in comparison with other techniques.

Pengzhen Ren, Yun Xiao, Xiaojun Chang et al.

Active learning (AL) attempts to maximize the performance gain of the model by marking the fewest samples. Deep learning (DL) is greedy for data and requires a large amount of data supply to optimize massive parameters, so that the model learns how to extract high-quality features. In recent years, due to the rapid development of internet technology, we are in an era of information torrents and we have massive amounts of data. In this way, DL has aroused strong interest of researchers and has been rapidly developed. Compared with DL, researchers have relatively low interest in AL. This is mainly because before the rise of DL, traditional machine learning requires relatively few labeled samples. Therefore, early AL is difficult to reflect the value it deserves. Although DL has made breakthroughs in various fields, most of this success is due to the publicity of the large number of existing annotation datasets. However, the acquisition of a large number of high-quality annotated datasets consumes a lot of manpower, which is not allowed in some fields that require high expertise, especially in the fields of speech recognition, information extraction, medical images, etc. Therefore, AL has gradually received due attention. A natural idea is whether AL can be used to reduce the cost of sample annotations, while retaining the powerful learning capabilities of DL. Therefore, deep active learning (DAL) has emerged. Although the related research has been quite abundant, it lacks a comprehensive survey of DAL. This article is to fill this gap, we provide a formal classification method for the existing work, and a comprehensive and systematic overview. In addition, we also analyzed and summarized the development of DAL from the perspective of application. Finally, we discussed the confusion and problems in DAL, and gave some possible development directions for DAL.

Pengzhen Ren, Yun Xiao, Xiaojun Chang et al.

Deep learning has made breakthroughs and substantial in many fields due to its powerful automatic representation capabilities. It has been proven that neural architecture design is crucial to the feature representation of data and the final performance. However, the design of the neural architecture heavily relies on the researchers' prior knowledge and experience. And due to the limitations of human' inherent knowledge, it is difficult for people to jump out of their original thinking paradigm and design an optimal model. Therefore, an intuitive idea would be to reduce human intervention as much as possible and let the algorithm automatically design the neural architecture. Neural Architecture Search (NAS) is just such a revolutionary algorithm, and the related research work is complicated and rich. Therefore, a comprehensive and systematic survey on the NAS is essential. Previously related surveys have begun to classify existing work mainly based on the key components of NAS: search space, search strategy, and evaluation strategy. While this classification method is more intuitive, it is difficult for readers to grasp the challenges and the landmark work involved. Therefore, in this survey, we provide a new perspective: beginning with an overview of the characteristics of the earliest NAS algorithms, summarizing the problems in these early NAS algorithms, and then providing solutions for subsequent related research work. Besides, we conduct a detailed and comprehensive analysis, comparison, and summary of these works. Finally, we provide some possible future research directions.

Lei Zhu, Zi Huang, Zhihui Li et al.

Unsupervised hashing can desirably support scalable content-based image retrieval (SCBIR) for its appealing advantages of semantic label independence, memory and search efficiency. However, the learned hash codes are embedded with limited discriminative semantics due to the intrinsic limitation of image representation. To address the problem, in this paper, we propose a novel hashing approach, dubbed as \emph{Discrete Semantic Transfer Hashing} (DSTH). The key idea is to \emph{directly} augment the semantics of discrete image hash codes by exploring auxiliary contextual modalities. To this end, a unified hashing framework is formulated to simultaneously preserve visual similarities of images and perform semantic transfer from contextual modalities. Further, to guarantee direct semantic transfer and avoid information loss, we explicitly impose the discrete constraint, bit--uncorrelation constraint and bit-balance constraint on hash codes. A novel and effective discrete optimization method based on augmented Lagrangian multiplier is developed to iteratively solve the optimization problem. The whole learning process has linear computation complexity and desirable scalability. Experiments on three benchmark datasets demonstrate the superiority of DSTH compared with several state-of-the-art approaches.

De Cheng, Yihong Gong, Zhihui Li et al.

Learning the distance metric between pairs of examples is of great importance for visual recognition, especially for person re-identification (Re-Id). Recently, the contrastive and triplet loss are proposed to enhance the discriminative power of the deeply learned features, and have achieved remarkable success. As can be seen, either the contrastive or triplet loss is just one special case of the Euclidean distance relationships among these training samples. Therefore, we propose a structured graph Laplacian embedding algorithm, which can formulate all these structured distance relationships into the graph Laplacian form. The proposed method can take full advantages of the structured distance relationships among these training samples, with the constructed complete graph. Besides, this formulation makes our method easy-to-implement and super-effective. When embedding the proposed algorithm with the softmax loss for the CNN training, our method can obtain much more robust and discriminative deep features with inter-personal dispersion and intra-personal compactness, which is essential to person Re-Id. We illustrate the effectiveness of our proposed method on top of three popular networks, namely AlexNet, DGDNet and ResNet50, on recent four widely used Re-Id benchmark datasets. Our proposed method achieves state-of-the-art performances.

Minnan Luo, Xiaojun Chang, Zhihui Li et al.

The heterogeneity-gap between different modalities brings a significant challenge to multimedia information retrieval. Some studies formalize the cross-modal retrieval tasks as a ranking problem and learn a shared multi-modal embedding space to measure the cross-modality similarity. However, previous methods often establish the shared embedding space based on linear mapping functions which might not be sophisticated enough to reveal more complicated inter-modal correspondences. Additionally, current studies assume that the rankings are of equal importance, and thus all rankings are used simultaneously, or a small number of rankings are selected randomly to train the embedding space at each iteration. Such strategies, however, always suffer from outliers as well as reduced generalization capability due to their lack of insightful understanding of procedure of human cognition. In this paper, we involve the self-paced learning theory with diversity into the cross-modal learning to rank and learn an optimal multi-modal embedding space based on non-linear mapping functions. This strategy enhances the model's robustness to outliers and achieves better generalization via training the model gradually from easy rankings by diverse queries to more complex ones. An efficient alternative algorithm is exploited to solve the proposed challenging problem with fast convergence in practice. Extensive experimental results on several benchmark datasets indicate that the proposed method achieves significant improvements over the state-of-the-arts in this literature.

Yun Song, Zhihui Li

There are several methods for constructing secret sharing schemes, one of which is based on coding theory. Theoretically, every linear code can be used to construct secret sharing schemes. However, in general, determining the access structures of the schemes based on linear codes is very hard. This paper proposed the concept of minimal linear code, which makes the determination of the access structures of the schemes based on the duals of minimal linear codes easier. It is proved that the shortening codes of minimal linear codes are also minimal ones. Then the conditions whether several types of irreducible cyclic codes are minimal linear codes are presented. Furthermore, the access structures of secret sharing schemes based on the duals of minimal linear codes are studied, and these access structures in specific examples are obtained through programming.

Yun Song, Zhihui Li

Irreducible cyclic codes are one of the largest known classes of block codes which have been investigated for a long time. However, their weight distributions are known only for a few cases. In this paper, a class of irreducible cyclic codes are studied and their weight distributions are determined. Moreover, all codewords of some irreducible cyclic codes are obtained through programming in order to explain their distributions. The number of distinct nonzero weights in these codes dealt with in this paper varies among 1,2,3,6,8.