Ying Huang

Lingyu Xiong, Jinjin Shi, Xuran Xu et al.

Vision Foundation Models (VFMs) have demonstrated impressive representational capabilities. However, adapting them to downstream tasks via full fine-tuning incurs prohibitive computational and storage overhead. Parameter-Efficient Fine-Tuning (PEFT) has emerged as a compelling alternative, aiming to achieve performance parity with full fine-tuning at minimal training costs. Nonetheless, applying PEFT to VFMs for dense prediction tasks remains challenging due to the structural and distributional gaps. To bridge these gaps, we propose \textbf{S}cale-\textbf{I}ntegrated \textbf{G}lobal \textbf{M}odulation \textbf{A}dapter (\textbf{SIGMA}), a novel lightweight PEFT method, which consists of two modules: scale-adaptive fusion and semantic modulation. Specifically, the scale-adaptive fusion module is utilized to bridge structural gaps by enhancing the extraction of multi-granularity visual information. Furthermore, SIGMA introduces semantic modulation on the fusion features to perform global feature alignment to further eliminate the distribution gap. This design facilitates unified spatial and distributional adaptation, requiring only 1.72\% trainable parameters relative to the VFM backbone. Comprehensive experiments across various downstream dense tasks and multiple VFM backbones demonstrate that SIGMA achieves consistent and superior performance over state-of-the-art PEFT methods.

Si Shen, Jiangfeng Liu, Litao Lin et al.

The academic literature of social sciences records human civilization and studies human social problems. With its large-scale growth, the ways to quickly find existing research on relevant issues have become an urgent demand for researchers. Previous studies, such as SciBERT, have shown that pre-training using domain-specific texts can improve the performance of natural language processing tasks. However, the pre-trained language model for social sciences is not available so far. In light of this, the present research proposes a pre-trained model based on the abstracts published in the Social Science Citation Index (SSCI) journals. The models, which are available on GitHub (https://github.com/S-T-Full-Text-Knowledge-Mining/SSCI-BERT), show excellent performance on discipline classification, abstract structure-function recognition, and named entity recognition tasks with the social sciences literature.

Jiaqing Liang, Jinyi Han, Weijia Li et al.

Long-horizon large language model (LLM) agents are fundamentally limited by context. As interactions become longer, tool descriptions, retrieved memories, and raw environmental feedback accumulate and push out the information needed for decision-making. At the same time, useful experience gained from tasks is often lost across episodes. We argue that long-horizon performance is determined not by context length, but by how much decision-relevant information is maintained within a finite context budget. We present GenericAgent (GA), a general-purpose, self-evolving LLM agent system built around a single principle: context information density maximization. GA implements this through four closely connected components: a minimal atomic tool set that keeps the interface simple, a hierarchical on-demand memory that only shows a small high-level view by default, a self-evolution mechanism that turns verified past trajectories into reusable SOPs and executable code, and a context truncation and compression layer that maintains information density during long executions. Across task completion, tool use efficiency, memory effectiveness, self-evolution, and web browsing, GA consistently outperforms leading agent systems while using significantly fewer tokens and interactions, and it continues to evolve over time. Project: https://github.com/lsdefine/GenericAgent

Shaokang Wang, Pei Fu, Ruoceng Zhang et al.

While Large Vision-Language Models (LVLMs) have significantly advanced GUI agents' capabilities in parsing textual instructions, interpreting screen content, and executing tasks, a critical challenge persists: the irreversibility of agent operations, where a single erroneous action can trigger catastrophic deviations. To address this, we propose the GUI Action Critic's Data Flywheel System (GAIA), a training framework that enables the models to have iterative critic capabilities, which are used to improve the Test-Time Scaling (TTS) of basic GUI agents' performance. Specifically, we train an Intuitive Critic Model (ICM) using positive and negative action examples from a base agent first. This critic evaluates the immediate correctness of the agent's intended actions, thereby selecting operations with higher success probability. Then, the initial critic guides agent actions to collect refined positive/negative samples, initiating the self-improving cycle. The augmented data then trains a second-round critic with enhanced discernment capability. We conduct experiments on various datasets and demonstrate that the proposed ICM can improve the test-time performance of various closed-source and open-source models, and the performance can be gradually improved as the data is recycled. The code and dataset will be publicly released.

Yunlong Wang, Jinjin Shi, Wenbin Gao et al.

Text-to-image diffusion models often face a severe trilemma in human portrait generation: text-image alignment, photorealism, and human-perceived aesthetics inherently inhibit one another. Supervised Fine-Tuning (SFT) is an effective method for enhancing the photorealism of image generation. However, it often leads to overfitting to the training dataset, corrupts pre-trained image priors, and degrades alignment or aesthetics. To break this bottleneck, we propose a feature supervision paradigm for Multimodal Diffusion Transformers (MM-DiT). Specifically, we introduce a lightweight cross-modal alignment mechanism that implicitly extracts multi-granularity vision-aligned text representations from SigLIP 2 and applies supervision to the image branch of MM-DiT during the training stage, with zero extra inference overhead. Our method injects vision-aligned text guidance while preserving the base model's original generalization, avoiding degradation caused by SFT. Furthermore, our method directly mines implicit multi-granularity aesthetic signals from pre-trained vision foundation models to optimize human-perceived aesthetics. Extensive experiments on MM-DiTs show that our method pushes the Pareto frontier and achieves synergistic improvements across text-image alignment, photorealism, and human-perceived aesthetics.

Beining Wu, Fuyou Mao, Jiong Lin et al.

Generative engines (GEs) are reshaping information access by replacing ranked links with citation-grounded answers, yet current Generative Engine Optimization (GEO) methods optimize each instance in isolation, unable to accumulate or transfer effective strategies across tasks and engines. We reframe GEO as a strategy learning problem and propose MAGEO, a multi-agent framework in which coordinated planning, editing, and fidelity-aware evaluation serve as the execution layer, while validated editing patterns are progressively distilled into reusable, engine-specific optimization skills. To enable controlled assessment, we introduce a Twin Branch Evaluation Protocol for causal attribution of content edits and DSV-CF, a dual-axis metric that unifies semantic visibility with attribution accuracy. We further release MSME-GEO-Bench, a multi-scenario, multi-engine benchmark grounded in real-world queries. Experiments on three mainstream engines show that MAGEO substantially outperforms heuristic baselines in both visibility and citation fidelity, with ablations confirming that engine-specific preference modeling and strategy reuse are central to these gains, suggesting a scalable learning-driven paradigm for trustworthy GEO. Code is available at https://github.com/Wu-beining/MAGEO

Adrian de Luis, Minh Tran, Taisei Hanyu et al.

As climate change intensifies, the global imperative to shift towards sustainable energy sources becomes more pronounced. Photovoltaic (PV) energy is a favored choice due to its reliability and ease of installation. Accurate mapping of PV installations is crucial for understanding their adoption and informing energy policy. To meet this need, we introduce the SolarFormer, designed to segment solar panels from aerial imagery, offering insights into their location and size. However, solar panel identification in Computer Vision is intricate due to various factors like weather conditions, roof conditions, and Ground Sampling Distance (GSD) variations. To tackle these complexities, we present the SolarFormer, featuring a multi-scale Transformer encoder and a masked-attention Transformer decoder. Our model leverages low-level features and incorporates an instance query mechanism to enhance the localization of solar PV installations. We rigorously evaluated our SolarFormer using diverse datasets, including GGE (France), IGN (France), and USGS (California, USA), across different GSDs. Our extensive experiments consistently demonstrate that our model either matches or surpasses state-of-the-art models, promising enhanced solar panel segmentation for global sustainable energy initiatives.

Fei Ge, Ying Huang, Jie Liu et al.

Existing deep image watermarking methods follow a fixed embedding-distortion-extraction pipeline, where the embedder and extractor are weakly coupled through a final loss and optimized in isolation. This design lacks explicit collaboration, leaving no structured mechanism for the embedder to incorporate decoding-aware cues or for the extractor to guide embedding during training. To address this architectural limitation, we rethink deep image watermarking by reformulating embedding and extraction as explicitly collaborative components. To realize this reformulation, we introduce a Collaborative Interaction Mechanism (CIM) that establishes direct, bidirectional communication between the embedder and extractor, enabling a mutual-teacher training paradigm and coordinated optimization. Built upon this explicitly collaborative architecture, we further propose an Adaptive Feature Modulation Module (AFMM) to support effective interaction. AFMM enables content-aware feature regulation by decoupling modulation structure and strength, guiding watermark embedding toward stable image features while suppressing host interference during extraction. Under CIM, the AFMMs on both sides form a closed-loop collaboration that aligns embedding behavior with extraction objectives. This architecture-level redesign changes how robustness is learned in watermarking systems. Rather than relying on exhaustive distortion simulation, robustness emerges from coordinated representation learning between embedding and extraction. Experiments on real-world and AI-generated datasets demonstrate that the proposed method consistently outperforms state-of-the-art approaches in watermark extraction accuracy while maintaining high perceptual quality, showing strong robustness and generalization.

Qifei Jia, Yu Liu, Yajie Chai et al.

Instruction-based image editing has garnered significant attention due to its direct interaction with users. However, real-world user instructions are immensely diverse, and existing methods often fail to generalize effectively to instructions outside their training domain, limiting their practical application. To address this, we propose Lego-Edit, which leverages the generalization capability of Multi-modal Large Language Model (MLLM) to organize a suite of model-level editing tools to tackle this challenge. Lego-Edit incorporates two key designs: (1) a model-level toolkit comprising diverse models efficiently trained on limited data and several image manipulation functions, enabling fine-grained composition of editing actions by the MLLM; and (2) a three-stage progressive reinforcement learning approach that uses feedback on unannotated, open-domain instructions to train the MLLM, equipping it with generalized reasoning capabilities for handling real-world instructions. Experiments demonstrate that Lego-Edit achieves state-of-the-art performance on GEdit-Bench and ImgBench. It exhibits robust reasoning capabilities for open-domain instructions and can utilize newly introduced editing tools without additional fine-tuning. Code is available: https://github.com/xiaomi-research/lego-edit.

Brian D. Williamson, Drew King, Ying Huang

Estimating a prediction function is a fundamental component of many data analyses. The super learner ensemble, a particular implementation of stacking, has desirable theoretical properties and has been used successfully in many applications. Dimension reduction can be accomplished by using variable screening algorithms (screeners), including the lasso, within the ensemble prior to fitting other prediction algorithms. However, the performance of a super learner using the lasso for dimension reduction has not been fully explored in cases where the lasso is known to perform poorly. We provide empirical results that suggest that a diverse set of candidate screeners should be used to protect against poor performance of any one screener, similar to the guidance for choosing a library of prediction algorithms for the super learner. These results are further illustrated through the analysis of HIV-1 antibody data.

Wenda Liu, Zhigang Song, Shuai Nie et al.

LLM-based universal information extraction (UIE) methods often rely on additional information beyond the original training data, which increases training complexity yet often yields limited gains. To address this, we propose ProUIE, a Macro-to-Micro progressive learning approach that improves UIE without introducing any external information. ProUIE consists of three stages: (i) macro-level Complete Modeling (CM), which learns NER, RE, and EE along their intrinsic difficulty order on the full training data to build a unified extraction foundation, (ii) meso-level Streamlined Alignment (SA), which operates on sampled data with simplified target formats, streamlining and regularizing structured outputs to make them more concise and controllable, and (iii) micro-level Deep Exploration (DE), which applies GRPO with stepwise fine-grained rewards (SFR) over structural units to guide exploration and improve performance. Experiments on 36 public datasets show that ProUIE consistently improves unified extraction, outperforming strong instruction-tuned baselines on average for NER and RE while using a smaller backbone, and it further demonstrates clear gains in large-scale production-oriented information extraction.

Yang Li, Yuchen Liu, Haoyu Lu et al.

Recent progress in Multimodal Large Language Models (MLLMs) has enabled mobile GUI agents capable of visual perception, cross-modal reasoning, and interactive control. However, existing benchmarks are largely English-centric and fail to capture the linguistic and interaction characteristics of the Chinese mobile ecosystem. They also focus on isolated skills such as GUI grounding or offline agent, lacking a unified and fine-grained framework to assess the full capability chain from perception to execution. To address this gap, we introduce GUI-CEval, the first comprehensive benchmark for Chinese mobile GUI agents, built entirely on physical device environments. GUI-CEval spans 201 mainstream apps across four device types and adopts a two-level structure that evaluates both atomic abilities and realistic application-level performance along five dimensions: perception, planning, reflection, execution, and evaluation. All data are collected and verified through multi-stage manual processes to ensure authenticity and reproducibility. Extensive experiments on 20 representative MLLMs and multi-agent systems show that while models such as Qwen2.5-VL and UI-TARS perform competitively, most MLLMs still exhibit clear weaknesses in reflective decision-making and post-action self-evaluation, limiting their reliability in real-world interactions. We hope GUI-CEval provides a comprehensive and interpretable benchmark to guide capability diagnosis and advance the development of Chinese mobile GUI agents.

Changpeng Yang, Jinyang Wu, Yuchen Liu et al.

Reinforcement learning has emerged as a paradigm for post-training large language models, boosting their reasoning capabilities. Such approaches compute an advantage value for each sample, reflecting better or worse performance than expected, thereby yielding both positive and negative signals for training. However, the indiscriminate mixing of the two signals in existing methods, especially from the early stages, may lead to ambiguous guidance and limited gains. To address this issue, we propose **CAPO** (**C**urriculum **A**dvantage **P**olicy **O**ptimization), an adaptive curriculum mechanism based on advantage signals. The proposed mechanism bootstraps imitation learning with positive-only advantage samples to establish robust foundations, and subsequently introduces negative signals to cultivate discriminative capabilities, thereby improving generalization across complex scenarios. Compatible with diverse optimization methods including GRPO, PPO, RLOO, and Reinforce++, our method consistently achieves stable and significant improvements in mathematical reasoning tasks, and further generalizes effectively to multimodal Graphical User Interface (GUI) reasoning scenarios, establishing itself as a versatile and robust optimization framework.

Jinyi Han, Ying Huang, Ying Liao et al.

Large Reasoning Models (LRMs) have achieved impressive performance on challenging tasks, yet their deep reasoning often incurs substantial computational costs. To achieve efficient reasoning, existing reinforcement learning methods still struggle to construct short reasoning path during the rollout stage, limiting effective learning. Inspired by Evidence Accumulation Models, we find that LRMs have accumulated sufficient information early in reasoning, making further reasoning steps redundant. Based on this insight, we propose Just-Enough Thinking (JET), which trains models to proactively terminate unnecessary reasoning. JET performs trajectory truncation during rollout to expose the model to short, distributionally consistent reasoning paths. Besides, it uses a quality-controlled length reward to better encourage concise reasoning while maintaining correctness. Extensive experiments demonstrate that JET significantly improves reasoning efficiency without sacrificing accuracy. Especially, DeepSeek-Distill-Qwen-1.5B achieves a 4.6% accuracy gain while reducing output length by 46.3% on the Olympiad benchmark. Our code is available in the GitHub.

Songtao Wei, Yi Li, Bohan Zhang et al.

Parameter-Efficient Fine-Tuning (PEFT) has emerged as a practical paradigm for adapting large language models (LLMs) without updating all parameters. Most existing approaches, such as LoRA and PiSSA, rely on low-rank decompositions of weight updates. However, the low-rank assumption may restrict expressivity, particularly in task-specific adaptation scenarios where singular values are distributed relatively uniformly. To address this limitation, we propose CoSA (Compressed Sensing-Based Adaptation), a new PEFT method extended from compressed sensing theory. Instead of constraining weight updates to a low-rank subspace, CoSA expresses them through fixed random projection matrices and a compact learnable core. We provide a formal theoretical analysis of CoSA as a synthesis process, proving that weight updates can be compactly encoded into a low-dimensional space and mapped back through random projections. Extensive experimental results show that CoSA provides a principled perspective for efficient and expressive multi-scale model adaptation. Specifically, we evaluate CoSA on 10 diverse tasks, including natural language understanding and generation, employing 5 models of different scales from RoBERTa, Llama, and Qwen families. Across these settings, CoSA consistently matches or outperforms state-of-the-art PEFT methods.

Dezhi Kong, Zhengzhao Feng, Qiliang Liang et al.

Multimodal large language models (MLLMs) have made significant progress in mobile agent development, yet their capabilities are predominantly confined to a reactive paradigm, where they merely execute explicit user commands. The emerging paradigm of proactive intelligence, where agents autonomously anticipate needs and initiate actions, represents the next frontier for mobile agents. However, its development is critically bottlenecked by the lack of benchmarks that can address real-world complexity and enable objective, executable evaluation. To overcome these challenges, we introduce ProactiveMobile, a comprehensive benchmark designed to systematically advance research in this domain. ProactiveMobile formalizes the proactive task as inferring latent user intent across four dimensions of on-device contextual signals and generating an executable function sequence from a comprehensive function pool of 63 APIs. The benchmark features over 3,660 instances of 14 scenarios that embrace real-world complexity through multi-answer annotations. To ensure quality, a team of 30 experts conducts a final audit of the benchmark, verifying factual accuracy, logical consistency, and action feasibility, and correcting any non-compliant entries. Extensive experiments demonstrate that our fine-tuned Qwen2.5-VL-7B-Instruct achieves a success rate of 19.15%, outperforming o1 (15.71%) and GPT-5 (7.39%). This result indicates that proactivity is a critical competency widely lacking in current MLLMs, yet it is learnable, emphasizing the importance of the proposed benchmark for proactivity evaluation.

Shaojie Zhang, Pei Fu, Ruoceng Zhang et al.

Autonomous Graphical User Interface (GUI) agents rely on accurate GUI grounding, which maps language instructions to on-screen coordinates, to execute user commands. However, current models, whether trained via supervised fine-tuning (SFT) or reinforcement fine-tuning (RFT), lack self-awareness of their capability boundaries, leading to overconfidence and unreliable predictions. We first systematically evaluate probabilistic and verbalized confidence in general and GUI-specific models, revealing a misalignment between confidence and actual accuracy, which is particularly critical in dynamic GUI automation tasks, where single errors can cause task failure. To address this, we propose HyperClick, a novel framework that enhances reliable GUI grounding through uncertainty calibration. HyperClick introduces a dual reward mechanism, combining a binary reward for correct actions with a truncated Gaussian-based spatial confidence modeling, calibrated using the Brier score. This approach jointly optimizes grounding accuracy and confidence reliability, fostering introspective self-criticism. Extensive experiments on seven challenge benchmarks show that HyperClick achieves state-of-the-art performance while providing well-calibrated confidence. By enabling explicit confidence calibration and introspective self-criticism, HyperClick reduces overconfidence and supports more reliable GUI automation.

HyperAI Team, Yuchen Liu, Kaiyang Han et al.

Current multimodal large lanauge models possess strong perceptual and reasoning capabilities, however high computational and memory requirements make them difficult to deploy directly on on-device environments. While small-parameter models are progressively endowed with strong general capabilities, standard Vision Transformer (ViT) encoders remain a critical bottleneck, suffering from excessive latency and memory consumption when processing high-resolution inputs.To address these challenges, we introduce HyperVL, an efficient multimodal large language model tailored for on-device inference. HyperVL adopts an image-tiling strategy to cap peak memory usage and incorporates two novel techniques: (1) a Visual Resolution Compressor (VRC) that adaptively predicts optimal encoding resolutions to eliminate redundant computation, and (2) Dual Consistency Learning (DCL), which aligns multi-scale ViT encoders within a unified framework, enabling dynamic switching between visual branches under a shared LLM. Extensive experiments demonstrate that HyperVL achieves state-of-the-art performance among models of comparable size across multiple benchmarks. Furthermore, it significantly significantly reduces latency and power consumption on real mobile devices, demonstrating its practicality for on-device multimodal inference.

Shaojie Zhang, Ruoceng Zhang, Pei Fu et al.

In the field of AI-driven human-GUI interaction automation, while rapid advances in multimodal large language models and reinforcement fine-tuning techniques have yielded remarkable progress, a fundamental challenge persists: their interaction logic significantly deviates from natural human-GUI communication patterns. To fill this gap, we propose "Blink-Think-Link" (BTL), a brain-inspired framework for human-GUI interaction that mimics the human cognitive process between users and graphical interfaces. The system decomposes interactions into three biologically plausible phases: (1) Blink - rapid detection and attention to relevant screen areas, analogous to saccadic eye movements; (2) Think - higher-level reasoning and decision-making, mirroring cognitive planning; and (3) Link - generation of executable commands for precise motor control, emulating human action selection mechanisms. Additionally, we introduce two key technical innovations for the BTL framework: (1) Blink Data Generation - an automated annotation pipeline specifically optimized for blink data, and (2) BTL Reward -- the first rule-based reward mechanism that enables reinforcement learning driven by both process and outcome. Building upon this framework, we develop a GUI agent model named BTL-UI, which demonstrates competitive performance across both static GUI understanding and dynamic interaction tasks in comprehensive benchmarks. These results provide conclusive empirical validation of the framework's efficacy in developing advanced GUI Agents.

Yuanbin Man, Ying Huang, Chengming Zhang et al.

The advancements in large language models (LLMs) have propelled the improvement of video understanding tasks by incorporating LLMs with visual models. However, most existing LLM-based models (e.g., VideoLLaMA, VideoChat) are constrained to processing short-duration videos. Recent attempts to understand long-term videos by extracting and compressing visual features into a fixed memory size. Nevertheless, those methods leverage only visual modality to merge video tokens and overlook the correlation between visual and textual queries, leading to difficulties in effectively handling complex question-answering tasks. To address the challenges of long videos and complex prompts, we propose AdaCM$^2$, which, for the first time, introduces an adaptive cross-modality memory reduction approach to video-text alignment in an auto-regressive manner on video streams. Our extensive experiments on various video understanding tasks, such as video captioning, video question answering, and video classification, demonstrate that AdaCM$^2$ achieves state-of-the-art performance across multiple datasets while significantly reducing memory usage. Notably, it achieves a 4.5% improvement across multiple tasks in the LVU dataset with a GPU memory consumption reduction of up to 65%.

Ying Huang, Yuanbin Man, Wenqi Jia et al.

Adapter-based fine-tuning has gained remarkable attention in adapting large pre-trained vision language models (VLMs) for a wide range of downstream tasks efficiently. In this paradigm, only the inserted adapters are fine-tuned, without the need for training the original VLM backbone. Existing works scale adapters by integrating them into every layer of VLMs to increase the capacity of adapters. However, these methods face two primary limitations: 1) limited compression rate due to ignoring cross-layer redundancy, and 2) limited representational capacity across homogeneous adapters. In this paper, we propose a novel vision-language fine-tuning framework based on cross-layer tensor ring decomposition (TRD) with the integration and collaboration of diverse adapters, called AdaRing, achieving ultra-light parameter-efficient adaptation of VLMs on various tasks. To remove the high redundancy that exists among adapters across layers, we exploit the tensor-level low-rankness to formulate adapters as layer-shared tensor cores and layer-specific slices. Moreover, guided by generalization-aware fine-tuning, diverse rank-driven adapters cooperate to handle tasks that require different representations. Our experiments show that the proposed AdaRing achieves the state-of-the-art performance while reducing average training parameters by 90%.

Jianning Chi, Zelan Li, Huixuan Wu et al.

Weakly-supervised methods typically guided the pixel-wise training by comparing the predictions to single-level labels containing diverse segmentation-related information at once, but struggled to represent delicate feature differences between nodule and background regions and confused incorrect information, resulting in underfitting or overfitting in the segmentation predictions. In this work, we propose a weakly-supervised network that generates multi-level labels from four-point annotation to refine diverse constraints for delicate nodule segmentation. The Distance-Similarity Fusion Prior referring to the points annotations filters out information irrelevant to nodules. The bounding box and pure foreground/background labels, generated from the point annotation, guarantee the rationality of the prediction in the arrangement of target localization and the spatial distribution of target/background regions, respectively. Our proposed network outperforms existing weakly-supervised methods on two public datasets with respect to the accuracy and robustness, improving the applicability of deep-learning based segmentation in the clinical practice of thyroid nodule diagnosis.

Yuanbin Man, Ying Huang, Zhile Ren et al.

4D generation, or dynamic 3D content generation, integrates spatial, temporal, and view dimensions to model realistic dynamic scenes, playing a foundational role in advancing world models and physical AI. However, maintaining long-chain consistency across both frames and viewpoints through the unique spatio-camera-motion (SCM) attention mechanism introduces substantial computational and memory overhead, often leading to out-of-memory (OOM) failures and prohibitive generation times. To address these challenges, we propose Turbo4DGen, an ultra-fast acceleration framework for diffusion-based multi-view 4D content generation. Turbo4DGen introduces a spatiotemporal cache mechanism that persistently reuses intermediate attention across denoising steps, combined with dynamically semantic-aware attention pruning and an adaptive SCM chain bypass scheduler, to drastically reduce redundant SCM attention computation. Our experimental results show that Turbo4DGen achieves an average 9.7$\times$ speedup without quality degradation on the ObjaverseDy and Consistent4D datasets. To the best of our knowledge, Turbo4DGen is the first dedicated acceleration framework for 4D generation.

Xiao Zhang, Zhuo Jin, Shaoxuan Wu et al.

Accurate vessel segmentation is crucial to assist in clinical diagnosis by medical experts. However, the intricate tree-like tubular structure of blood vessels poses significant challenges for existing segmentation algorithms. Small vascular branches are often overlooked due to their low contrast compared to surrounding tissues, leading to incomplete vessel segmentation. Furthermore, the complex vascular topology prevents the model from accurately capturing and reconstructing vascular structure, resulting in incorrect topology, such as breakpoints at the bifurcation of the vascular tree. To overcome these challenges, we propose a novel vessel segmentation framework called PASC Net. It includes two key modules: a plug-and-play shape self-learning convolutional (SSL) module that optimizes convolution kernel design, and a hierarchical topological constraint (HTC) module that ensures vascular connectivity through topological constraints. Specifically, the SSL module enhances adaptability to vascular structures by optimizing conventional convolutions into learnable strip convolutions, which improves the network's ability to perceive fine-grained features of tubular anatomies. Furthermore, to better preserve the coherence and integrity of vascular topology, the HTC module incorporates hierarchical topological constraints-spanning linear, planar, and volumetric levels-which serve to regularize the network's representation of vascular continuity and structural consistency. We replaced the standard convolutional layers in U-Net, FCN, U-Mamba, and nnUNet with SSL convolutions, leading to consistent performance improvements across all architectures. Furthermore, when integrated into the nnUNet framework, our method outperformed other methods on multiple metrics, achieving state-of-the-art vascular segmentation performance.

Minh Tran, Adrian De Luis, Haitao Liao et al.

As the impact of climate change escalates, the global necessity to transition to sustainable energy sources becomes increasingly evident. Renewable energies have emerged as a viable solution for users, with Photovoltaic energy being a favored choice for small installations due to its reliability and efficiency. Accurate mapping of PV installations is crucial for understanding the extension of its adoption and informing energy policy. To meet this need, we introduce S3Former, designed to segment solar panels from aerial imagery and provide size and location information critical for analyzing the impact of such installations on the grid. Solar panel identification is challenging due to factors such as varying weather conditions, roof characteristics, Ground Sampling Distance variations and lack of appropriate initialization weights for optimized training. To tackle these complexities, S3Former features a Masked Attention Mask Transformer incorporating a self-supervised learning pretrained backbone. Specifically, our model leverages low-level and high-level features extracted from the backbone and incorporates an instance query mechanism incorporated on the Transformer architecture to enhance the localization of solar PV installations. We introduce a self-supervised learning phase (pretext task) to improve the initialization weights on the backbone of S3Former. We evaluated S3Former using diverse datasets, demonstrate improvement state-of-the-art models.

Ying Huang, Liang Chen

In this note, we consider the problem of robust learning mixtures of linear regressions. We connect mixtures of linear regressions and mixtures of Gaussians with a simple thresholding, so that a quasi-polynomial time algorithm can be obtained under some mild separation condition. This algorithm has significantly better robustness than the previous result.

Chao Lv, Han Zhang, XinKai Du et al.

With the success of down streaming task using English pre-trained language model, the pre-trained Chinese language model is also necessary to get a better performance of Chinese NLP task. Unlike the English language, Chinese has its special characters such as glyph information. So in this article, we propose the Chinese pre-trained language model StyleBERT which incorporate the following embedding information to enhance the savvy of language model, such as word, pinyin, five stroke and chaizi. The experiments show that the model achieves well performances on a wide range of Chinese NLP tasks.

Zheng Li, Jingwen Ye, Mingli Song et al.

Existing state-of-the-art human pose estimation methods require heavy computational resources for accurate predictions. One promising technique to obtain an accurate yet lightweight pose estimator is knowledge distillation, which distills the pose knowledge from a powerful teacher model to a less-parameterized student model. However, existing pose distillation works rely on a heavy pre-trained estimator to perform knowledge transfer and require a complex two-stage learning procedure. In this work, we investigate a novel Online Knowledge Distillation framework by distilling Human Pose structure knowledge in a one-stage manner to guarantee the distillation efficiency, termed OKDHP. Specifically, OKDHP trains a single multi-branch network and acquires the predicted heatmaps from each, which are then assembled by a Feature Aggregation Unit (FAU) as the target heatmaps to teach each branch in reverse. Instead of simply averaging the heatmaps, FAU which consists of multiple parallel transformations with different receptive fields, leverages the multi-scale information, thus obtains target heatmaps with higher-quality. Specifically, the pixel-wise Kullback-Leibler (KL) divergence is utilized to minimize the discrepancy between the target heatmaps and the predicted ones, which enables the student network to learn the implicit keypoint relationship. Besides, an unbalanced OKDHP scheme is introduced to customize the student networks with different compression rates. The effectiveness of our approach is demonstrated by extensive experiments on two common benchmark datasets, MPII and COCO.

Chuang Liu, Shimin Yu, Ying Huang et al.

Link and sign prediction in complex networks bring great help to decision-making and recommender systems, such as in predicting potential relationships or relative status levels. Many previous studies focused on designing the special algorithms to perform either link prediction or sign prediction. In this work, we propose an effective model integration algorithm consisting of network embedding, network feature engineering, and an integrated classifier, which can perform the link and sign prediction in the same framework. Network embedding can accurately represent the characteristics of topological structures and cooperate with the powerful network feature engineering and integrated classifier can achieve better prediction. Experiments on several datasets show that the proposed model can achieve state-of-the-art or competitive performance for both link and sign prediction in spite of its generality. Interestingly, we find that using only very low network embedding dimension can generate high prediction performance, which can significantly reduce the computational overhead during training and prediction. This study offers a powerful methodology for multi-task prediction in complex networks.

Xinyu Dou, Cuijuan Liao, Hengqi Wang et al.

Nitrogen dioxide (NO2) is one of the most important atmospheric pollutants. However, current ground-level NO2 concentration data are lack of either high-resolution coverage or full coverage national wide, due to the poor quality of source data and the computing power of the models. To our knowledge, this study is the first to estimate the ground-level NO2 concentration in China with national coverage as well as relatively high spatiotemporal resolution (0.25 degree; daily intervals) over the newest past 6 years (2013-2018). We advanced a Random Forest model integrated K-means (RF-K) for the estimates with multi-source parameters. Besides meteorological parameters, satellite retrievals parameters, we also, for the first time, introduce socio-economic parameters to assess the impact by human activities. The results show that: (1) the RF-K model we developed shows better prediction performance than other models, with cross-validation R2 = 0.64 (MAPE = 34.78%). (2) The annual average concentration of NO2 in China showed a weak increasing trend . While in the economic zones such as Beijing-Tianjin-Hebei region, Yangtze River Delta, and Pearl River Delta, the NO2 concentration there even decreased or remained unchanged, especially in spring. Our dataset has verified that pollutant controlling targets have been achieved in these areas. With mapping daily nationwide ground-level NO2 concentrations, this study provides timely data with high quality for air quality management for China. We provide a universal model framework to quickly generate a timely national atmospheric pollutants concentration map with a high spatial-temporal resolution, based on improved machine learning methods.

Zheng Li, Ying Huang, Defang Chen et al.

Knowledge distillation is an effective method to transfer the knowledge from the cumbersome teacher model to the lightweight student model. Online knowledge distillation uses the ensembled prediction results of multiple student models as soft targets to train each student model. However, the homogenization problem will lead to difficulty in further improving model performance. In this work, we propose a new distillation method to enhance the diversity among multiple student models. We introduce Feature Fusion Module (FFM), which improves the performance of the attention mechanism in the network by integrating rich semantic information contained in the last block of multiple student models. Furthermore, we use the Classifier Diversification(CD) loss function to strengthen the differences between the student models and deliver a better ensemble result. Extensive experiments proved that our method significantly enhances the diversity among student models and brings better distillation performance. We evaluate our method on three image classification datasets: CIFAR-10/100 and CINIC-10. The results show that our method achieves state-of-the-art performance on these datasets.

Ying Huang, Shangfeng Qiu, Wenwei Zhang et al.

Researches using margin based comparison loss demonstrate the effectiveness of penalizing the distance between face feature and their corresponding class centers. Despite their popularity and excellent performance, they do not explicitly encourage the generic embedding learning for an open set recognition problem. In this paper, we analyse margin based softmax loss in probability view. With this perspective, we propose two general principles: 1) monotonic decreasing and 2) margin probability penalty, for designing new margin loss functions. Unlike methods optimized with single comparison metric, we provide a new perspective to treat open set face recognition as a problem of information transmission. And the generalization capability for face embedding is gained with more clean information. An auto-encoder architecture called Linear-Auto-TS-Encoder(LATSE) is proposed to corroborate this finding. Extensive experiments on several benchmarks demonstrate that LATSE help face embedding to gain more generalization capability and it boosted the single model performance with open training dataset to more than $99\%$ on MegaFace test.

Yan Yan, Ying Huang, Si Chen et al.

Recently, deep learning based facial expression recognition (FER) methods have attracted considerable attention and they usually require large-scale labelled training data. Nonetheless, the publicly available facial expression databases typically contain a small amount of labelled data. In this paper, to overcome the above issue, we propose a novel joint deep learning of facial expression synthesis and recognition method for effective FER. More specifically, the proposed method involves a two-stage learning procedure. Firstly, a facial expression synthesis generative adversarial network (FESGAN) is pre-trained to generate facial images with different facial expressions. To increase the diversity of the training images, FESGAN is elaborately designed to generate images with new identities from a prior distribution. Secondly, an expression recognition network is jointly learned with the pre-trained FESGAN in a unified framework. In particular, the classification loss computed from the recognition network is used to simultaneously optimize the performance of both the recognition network and the generator of FESGAN. Moreover, in order to alleviate the problem of data bias between the real images and the synthetic images, we propose an intra-class loss with a novel real data-guided back-propagation (RDBP) algorithm to reduce the intra-class variations of images from the same class, which can significantly improve the final performance. Extensive experimental results on public facial expression databases demonstrate the superiority of the proposed method compared with several state-of-the-art FER methods.

Ying Huang, Wenwei Zhang, Jinzhuo Wang

Face anti-spoofing is crucial for the security of face recognition system, by avoiding invaded with presentation attack. Previous works have shown the effectiveness of using depth and temporal supervision for this task. However, depth supervision is often considered only in a single frame, and temporal supervision is explored by utilizing certain signals which is not robust to the change of scenes. In this work, motivated by two stream ConvNets, we propose a novel two stream FreqSaptialTemporalNet for face anti-spoofing which simultaneously takes advantage of frequent, spatial and temporal information. Compared with existing methods which mine spoofing cues in multi-frame RGB image, we make multi-frame spectrum image as one input stream for the discriminative deep neural network, encouraging the primary difference between live and fake video to be automatically unearthed. Extensive experiments show promising improvement results using the proposed architecture. Meanwhile, we proposed a concise method to obtain a large amount of spoofing training data by utilizing a frequent augmentation pipeline, which contributes detail visualization between live and fake images as well as data insufficiency issue when training large networks.

Ying Huang, Jiankai Zhuang, Zengchang Qin

In multi-person pose estimation, the left/right joint type discrimination is always a hard problem because of the similar appearance. Traditionally, we solve this problem by stacking multiple refinement modules to increase network's receptive fields and capture more global context, which can also increase a great amount of computation. In this paper, we propose a Multi-level Network (MLN) that learns to aggregate features from lower-level (left/right information), upper-level (localization information), joint-limb level (complementary information) and global-level (context) information for discrimination of joint type. Through feature reuse and its intra-relation, MLN can attain comparable performance to other conventional methods while runtime speed retains at 42.2 FPS.

Ying Huang, Bin Sun, Haipeng Kan et al.

Human pose estimation has made significant advancement in recent years. However, the existing datasets are limited in their coverage of pose variety. In this paper, we introduce a novel benchmark FollowMeUp Sports that makes an important advance in terms of specific postures, self-occlusion and class balance, a contribution that we feel is required for future development in human body models. This comprehensive dataset was collected using an established taxonomy of over 200 standard workout activities with three different shot angles. The collected videos cover a wider variety of specific workout activities than previous datasets including push-up, squat and body moving near the ground with severe self-occlusion or occluded by some sport equipment and outfits. Given these rich images, we perform a detailed analysis of the leading human pose estimation approaches gaining insights for the success and failures of these methods.

Sayan Dasgupta, Ying Huang

Optimal biomarker combinations for treatment-selection can be derived by minimizing total burden to the population caused by the targeted disease and its treatment. However, when multiple biomarkers are present, including all in the model can be expensive and hurt model performance. To remedy this, we consider feature selection in optimization by minimizing an extended total burden that additionally incorporates biomarker measurement costs. Formulating it as a 0-norm penalized weighted classification, we develop various procedures for estimating linear and nonlinear combinations. Through simulations and a real data example, we demonstrate the importance of incorporating feature-selection and marker cost when deriving treatment-selection rules.

Ying Huang, Jiyang Dai, Chen Peng

A new method is developed to deal with the problem that a complex decentralized control system needs to keep centralized control performance. The systematic procedure emphasizes quickly finding the decentralized subcontrollers that matching the closed-loop performance and robustness characteristics of the centralized controller, which is featured by the fact that GA is used to optimize the design of centralized H-infinity controller K(s) and decentralized engine subcontroller KT(s), and that only one interface variable needs to satisfy decentralized control system requirement according to the proposed selection principle. The optimization design is motivated by the implementation issues where it is desirable to reduce the time in trial and error process and accurately find the best decentralized subcontrollers. The method is applied to decentralized control system design for a short takeoff and landing fighter. By comparing the simulation results of the decentralized control system with those of the centralized control system, the target of the decentralized control attains the performance and robustness of centralized control is validated.

Shuo Liu, Vijay John, Erik Blasch et al.

Context enhancement is critical for night vision (NV) applications, especially for the dark night situation without any artificial lights. In this paper, we present the infrared-to-visual (IR2VI) algorithm, a novel unsupervised thermal-to-visible image translation framework based on generative adversarial networks (GANs). IR2VI is able to learn the intrinsic characteristics from VI images and integrate them into IR images. Since the existing unsupervised GAN-based image translation approaches face several challenges, such as incorrect mapping and lack of fine details, we propose a structure connection module and a region-of-interest (ROI) focal loss method to address the current limitations. Experimental results show the superiority of the IR2VI algorithm over baseline methods.

Junxian He, Zhiting Hu, Taylor Berg-Kirkpatrick et al.

Correlated topic modeling has been limited to small model and problem sizes due to their high computational cost and poor scaling. In this paper, we propose a new model which learns compact topic embeddings and captures topic correlations through the closeness between the topic vectors. Our method enables efficient inference in the low-dimensional embedding space, reducing previous cubic or quadratic time complexity to linear w.r.t the topic size. We further speedup variational inference with a fast sampler to exploit sparsity of topic occurrence. Extensive experiments show that our approach is capable of handling model and data scales which are several orders of magnitude larger than existing correlation results, without sacrificing modeling quality by providing competitive or superior performance in document classification and retrieval.

Zizhao Zhang, Fuyong Xing, Hanzi Wang et al.

In this paper, we propose a simple but effective method for fast image segmentation. We re-examine the locality-preserving character of spectral clustering by constructing a graph over image regions with both global and local connections. Our novel approach to build graph connections relies on two key observations: 1) local region pairs that co-occur frequently will have a high probability to reside on a common object; 2) spatially distant regions in a common object often exhibit similar visual saliency, which implies their neighborship in a manifold. We present a novel energy function to efficiently conduct graph partitioning. Based on multiple high quality partitions, we show that the generated eigenvector histogram based representation can automatically drive effective unary potentials for a hierarchical random field model to produce multi-class segmentation. Sufficient experiments, on the BSDS500 benchmark, large-scale PASCAL VOC and COCO datasets, demonstrate the competitive segmentation accuracy and significantly improved efficiency of our proposed method compared with other state of the arts.

Junxian He, Ying Huang, Changfeng Liu et al.

A text network refers to a data type that each vertex is associated with a text document and the relationship between documents is represented by edges. The proliferation of text networks such as hyperlinked webpages and academic citation networks has led to an increasing demand for quickly developing a general sense of a new text network, namely text network exploration. In this paper, we address the problem of text network exploration through constructing a heterogeneous web of topics, which allows people to investigate a text network associating word level with document level. To achieve this, a probabilistic generative model for text and links is proposed, where three different relationships in the heterogeneous topic web are quantified. We also develop a prototype demo system named TopicAtlas to exhibit such heterogeneous topic web, and demonstrate how this system can facilitate the task of text network exploration. Extensive qualitative analyses are included to verify the effectiveness of this heterogeneous topic web. Besides, we validate our model on real-life text networks, showing that it preserves good performance on objective evaluation metrics.

Ying Huang, Hong Zheng, Haibin Ling et al.

Bird strikes present a huge risk for aircraft, especially since traditional airport bird surveillance is mainly dependent on inefficient human observation. Computer vision based technology has been proposed to automatically detect birds, determine bird flying trajectories, and predict aircraft takeoff delays. However, the characteristics of bird flight using imagery and the performance of existing methods applied to flying bird task are not well known. Therefore, we perform infrared flying bird tracking experiments using 12 state-of-the-art algorithms on a real BIRDSITE-IR dataset to obtain useful clues and recommend feature analysis. We also develop a Struck-scale method to demonstrate the effectiveness of multiple scale sampling adaption in handling the object of flying bird with varying shape and scale. The general analysis can be used to develop specialized bird tracking methods for airport safety, wildness and urban bird population studies.