arXiv subjects

Yu Zhou

Publications and source records attributed to Yu Zhou.

At least 127 records · Page 7Linked to original sources

arXiv2025

Task-Aware 3D Affordance Segmentation via 2D Guidance and Geometric Refinement

Understanding 3D scene-level affordances from natural language instructions is essential for enabling embodied agents to interact meaningfully in complex environments. However, this task remains challenging due to the need for semantic reasoning and spatial grounding. Existing methods mainly focus on object-level affordances or merely lift 2D predictions to 3D, neglecting rich geometric structure information in point clouds and incurring high computational costs. To address these limitations, we introduce Task-Aware 3D Scene-level Affordance segmentation (TASA), a novel geometry-optimized framework that jointly leverages 2D semantic cues and 3D geometric reasoning in a coarse-to-fine manner. To improve the affordance detection efficiency, TASA features a task-aware 2D affordance detection module to identify manipulable points from language and visual inputs, guiding the selection of task-relevant views. To fully exploit 3D geometric information, a 3D affordance refinement module is proposed to integrate 2D semantic priors with local 3D geometry, resulting in accurate and spatially coherent 3D affordance masks. Experiments on SceneFun3D demonstrate that TASA significantly outperforms the baselines in both accuracy and efficiency in scene-level affordance segmentation.

cs.CV↗

arXiv2025

Semibricks and wide subcategories in extended module categories

For $d\geq 1$, we define semibricks and wide subcategories in the $d$-extended hearts of bounded $t$-structures on a triangulated category. We show that these semibricks are in bijection with finite-length wide subcategories. When the $d$-extended heart is the $d$-extended module category $d\mbox{-}\mathrm{mod}Λ$ of a finite-dimensional algebra $Λ$ over a field, we define left/right-finite semibricks and left/right-finite wide subcategories in $d\mbox{-}\mathrm{mod}Λ$ and show bijections with $(d+1)$-term simple-minded collections, generalising the bijections between $2$-term simple-minded collections, left/right-finite wide subcategories and left/right-finite semibricks in $\mathrm{mod}Λ$. We use a relation between semibricks and silting complexes to characterise which mutations of $(d+1)$-term silting complexes are again $(d+1)$-term.

math.RT↗

arXiv2025

An Active Learning Pipeline for Biomedical Image Instance Segmentation with Minimal Human Intervention

Biomedical image segmentation is critical for precise structure delineation and downstream analysis. Traditional methods often struggle with noisy data, while deep learning models such as U-Net have set new benchmarks in segmentation performance. nnU-Net further automates model configuration, making it adaptable across datasets without extensive tuning. However, it requires a substantial amount of annotated data for cross-validation, posing a challenge when only raw images but no labels are available. Large foundation models offer zero-shot generalizability, but may underperform on specific datasets with unique characteristics, limiting their direct use for analysis. This work addresses these bottlenecks by proposing a data-centric AI workflow that leverages active learning and pseudo-labeling to combine the strengths of traditional neural networks and large foundation models while minimizing human intervention. The pipeline starts by generating pseudo-labels from a foundation model, which are then used for nnU-Net's self-configuration. Subsequently, a representative core-set is selected for minimal manual annotation, enabling effective fine-tuning of the nnU-Net model. This approach significantly reduces the need for manual annotations while maintaining competitive performance, providing an accessible solution for biomedical researchers to apply state-of-the-art AI techniques in their segmentation tasks. The code is available at https://github.com/MMV-Lab/AL_BioMed_img_seg.

cs.CV↗

arXiv2025

Non-Monotonicity in Fair Division of Graphs

We consider the problem of fairly allocating the vertices of a graph among $n$ agents, where the value of a bundle is determined by its cut value -- the number of edges with exactly one endpoint in the bundle. This model naturally captures applications such as team formation and network partitioning, where valuations are inherently non-monotonic: the marginal values may be positive, negative, or zero depending on the composition of the bundle. We focus on the fairness notion of envy-freeness up to one item (EF1) and explore its compatibility with several efficiency concepts such as Transfer Stability (TS) that prohibits single-item transfers that benefit one agent without making the other worse-off. For general graphs, our results uncover a non-monotonic relationship between the number of agents $n$ and the existence of allocations satisfying EF1 and transfer stability (TS): such allocations always exist for $n=2$, may fail to exist for $n=3$, but exist again for all $n\geq 4$. We further show that existence can be guaranteed for any $n$ by slightly weakening the efficiency requirement or by restricting the graph to forests. All of our positive results are achieved via efficient algorithms.

cs.GT↗

arXiv2025

MTL-KD: Multi-Task Learning Via Knowledge Distillation for Generalizable Neural Vehicle Routing Solver

Multi-Task Learning (MTL) in Neural Combinatorial Optimization (NCO) is a promising approach to train a unified model capable of solving multiple Vehicle Routing Problem (VRP) variants. However, existing Reinforcement Learning (RL)-based multi-task methods can only train light decoder models on small-scale problems, exhibiting limited generalization ability when solving large-scale problems. To overcome this limitation, this work introduces a novel multi-task learning method driven by knowledge distillation (MTL-KD), which enables the efficient training of heavy decoder models with strong generalization ability. The proposed MTL-KD method transfers policy knowledge from multiple distinct RL-based single-task models to a single heavy decoder model, facilitating label-free training and effectively improving the model's generalization ability across diverse tasks. In addition, we introduce a flexible inference strategy termed Random Reordering Re-Construction (R3C), which is specifically adapted for diverse VRP tasks and further boosts the performance of the multi-task model. Experimental results on 6 seen and 10 unseen VRP variants with up to 1000 nodes indicate that our proposed method consistently achieves superior performance on both uniform and real-world benchmarks, demonstrating robust generalization abilities.

cs.LG↗

arXiv2025

VidText: Towards Comprehensive Evaluation for Video Text Understanding

Visual texts embedded in videos carry rich semantic information, which is crucial for both holistic video understanding and fine-grained reasoning about local human actions. However, existing video understanding benchmarks largely overlook textual information, while OCR-specific benchmarks are constrained to static images, limiting their ability to capture the interaction between text and dynamic visual contexts. To address this gap, we propose VidText, a new benchmark designed for comprehensive and in-depth evaluation of video text understanding. VidText offers the following key features: 1) It covers a wide range of real-world scenarios and supports multilingual content, encompassing diverse settings where video text naturally appears. 2) It introduces a hierarchical evaluation framework with video-level, clip-level, and instance-level tasks, enabling assessment of both global summarization and local retrieval capabilities. 3) The benchmark also introduces a set of paired perception reasoning tasks, ranging from visual text perception to cross-modal reasoning between textual and visual information. Extensive experiments on 18 state-of-the-art Large Multimodal Models (LMMs) reveal that current models struggle across most tasks, with significant room for improvement. Further analysis highlights the impact of both model-intrinsic factors, such as input resolution and OCR capability, and external factors, including the use of auxiliary information and Chain-of-Thought reasoning strategies. We hope VidText will fill the current gap in video understanding benchmarks and serve as a foundation for future research on multimodal reasoning with video text in dynamic environments.

cs.CV↗

arXiv2025

Decorated Marked Surfaces with vortices: Cluster braid group vs. braid twist group

Let $\mathbf{S}$ be a marked surface with vortices (=punctures with extra $\mathbb{Z}_2$ symmetry). We study the decorated version $\mathbf{S}_\bigtriangleup$, where the $\mathbb{Z}_2$ symmetry lifts to the relation that the fourth power of the braid twist of any collision path (connecting a decoration in $\bigtriangleup$ and a vortex) is identity. We prove the following three groups are isomorphic: King-Qiu's cluster braid group associated to $\mathbf{S}$, the braid twist group of $\mathbf{S}_\bigtriangleup$ and the fundamental group of Bridgeland-Smith's moduli space of $\mathbf{S}$-framed GMN differentials. Moreover, we give finite presentations of such groups.

math.RT↗

arXiv2025

Josephson traveling-wave parametric amplifier based on low-intrinsic-loss coplanar lumped-element waveguide

We present a Josephson traveling-wave parametric amplifier (JTWPA) based on a low-loss coplanar lumped-element waveguide architecture. By employing open-stub capacitors and Manhattan-pattern junctions, our device achieves an insertion loss below 1~dB up to 12~GHz. We introduce windowed sinusoidal modulation for phase matching, demonstrating that a smooth transition in the impedance-modulation strength effectively suppresses intrinsic gain ripples. Using Tukey-windowed modulation with 8\% impedance variation, we achieve 20\text{--}23-dB~gain over 5-GHz bandwidth under ideal matching conditions. In a more practical circuit having impedance mismatches, the device maintains 17\text{--}20-dB gain over 4.8-GHz bandwidth with an added noise of 0.18~quanta above standard quantum limit at 20-dB gain and $-99$-dBm saturation power, while featuring zero to negative backward gain below the band-gap frequency.

physics.app-ph↗

arXiv2025

Distilling LLM Prior to Flow Model for Generalizable Agent's Imagination in Object Goal Navigation

The Object Goal Navigation (ObjectNav) task challenges agents to locate a specified object in an unseen environment by imagining unobserved regions of the scene. Prior approaches rely on deterministic and discriminative models to complete semantic maps, overlooking the inherent uncertainty in indoor layouts and limiting their ability to generalize to unseen environments. In this work, we propose GOAL, a generative flow-based framework that models the semantic distribution of indoor environments by bridging observed regions with LLM-enriched full-scene semantic maps. During training, spatial priors inferred from large language models (LLMs) are encoded as two-dimensional Gaussian fields and injected into target maps, distilling rich contextual knowledge into the flow model and enabling more generalizable completions. Extensive experiments demonstrate that GOAL achieves state-of-the-art performance on MP3D and Gibson, and shows strong generalization in transfer settings to HM3D. Codes and pretrained models are available at https://github.com/Badi-Li/GOAL.

cs.CV↗

arXiv2025

LLMs as Scalable, General-Purpose Simulators For Evolving Digital Agent Training

Digital agents require diverse, large-scale UI trajectories to generalize across real-world tasks, yet collecting such data is prohibitively expensive in both human annotation, infra and engineering perspectives. To this end, we introduce $\textbf{UI-Simulator}$, a scalable paradigm that generates structured UI states and transitions to synthesize training trajectories at scale. Our paradigm integrates a digital world simulator for diverse UI states, a guided rollout process for coherent exploration, and a trajectory wrapper that produces high-quality and diverse trajectories for agent training. We further propose $\textbf{UI-Simulator-Grow}$, a targeted scaling strategy that enables more rapid and data-efficient scaling by prioritizing high-impact tasks and synthesizes informative trajectory variants. Experiments on WebArena and AndroidWorld show that UI-Simulator rivals or surpasses open-source agents trained on real UIs with significantly better robustness, despite using weaker teacher models. Moreover, UI-Simulator-Grow matches the performance of Llama-3-70B-Instruct using only Llama-3-8B-Instruct as the base model, highlighting the potential of targeted synthesis scaling paradigm to continuously and efficiently enhance the digital agents.

cs.CL↗

arXiv2025

The Role of Video Generation in Enhancing Data-Limited Action Understanding

Video action understanding tasks in real-world scenarios always suffer data limitations. In this paper, we address the data-limited action understanding problem by bridging data scarcity. We propose a novel method that employs a text-to-video diffusion transformer to generate annotated data for model training. This paradigm enables the generation of realistic annotated data on an infinite scale without human intervention. We proposed the information enhancement strategy and the uncertainty-based label smoothing tailored to generate sample training. Through quantitative and qualitative analysis, we observed that real samples generally contain a richer level of information than generated samples. Based on this observation, the information enhancement strategy is proposed to enhance the informative content of the generated samples from two aspects: the environments and the characters. Furthermore, we observed that some low-quality generated samples might negatively affect model training. To address this, we devised the uncertainty-based label smoothing strategy to increase the smoothing of these samples, thus reducing their impact. We demonstrate the effectiveness of the proposed method on four datasets across five tasks and achieve state-of-the-art performance for zero-shot action recognition.

cs.CV↗

arXiv2025

When Semantics Mislead Vision: Mitigating Large Multimodal Models Hallucinations in Scene Text Spotting and Understanding

Large Multimodal Models (LMMs) have achieved impressive progress in visual perception and reasoning. However, when confronted with visually ambiguous or non-semantic scene text, they often struggle to accurately spot and understand the content, frequently generating semantically plausible yet visually incorrect answers, which we refer to as semantic hallucination. In this work, we investigate the underlying causes of semantic hallucination and identify a key finding: Transformer layers in LLM with stronger attention focus on scene text regions are less prone to producing semantic hallucinations. Thus, we propose a training-free semantic hallucination mitigation framework comprising two key components: (1) ZoomText, a coarse-to-fine strategy that identifies potential text regions without external detectors; and (2) Grounded Layer Correction, which adaptively leverages the internal representations from layers less prone to hallucination to guide decoding, correcting hallucinated outputs for non-semantic samples while preserving the semantics of meaningful ones. To enable rigorous evaluation, we introduce TextHalu-Bench, a benchmark of 1,740 samples spanning both semantic and non-semantic cases, with manually curated question answer pairs designed to probe model hallucinations. Extensive experiments demonstrate that our method not only effectively mitigates semantic hallucination but also achieves strong performance on public benchmarks for scene text spotting and understanding.

cs.CV↗

arXiv2025

Bin Packing and Covering: Pushing the Frontier on the Maximin Share Fairness

We study a fundamental fair allocation problem, where the agent's value is determined by the number of bins either used to pack or cover the items allocated to them. Fairness is evaluated using the maximin share (MMS) criterion. This problem is not only motivated by practical applications, but also serves as a natural framework for studying group fairness. As MMS is not always satisfiable, we consider two types of approximations: cardinal and ordinal. For cardinal approximation, we relax the requirements of being packed or covered for a bin, and for ordinal approximation, we relax the number of bins that are packed or covered. For all models of interest, we provide constant approximation algorithms.

cs.GT↗

arXiv2025

Towards More Accurate Diffusion Model Acceleration with A Timestep Tuner

A diffusion model, which is formulated to produce an image using thousands of denoising steps, usually suffers from a slow inference speed. Existing acceleration algorithms simplify the sampling by skipping most steps yet exhibit considerable performance degradation. By viewing the generation of diffusion models as a discretized integral process, we argue that the quality drop is partly caused by applying an inaccurate integral direction to a timestep interval. To rectify this issue, we propose a \textbf{timestep tuner} that helps find a more accurate integral direction for a particular interval at the minimum cost. Specifically, at each denoising step, we replace the original parameterization by conditioning the network on a new timestep, enforcing the sampling distribution towards the real one. Extensive experiments show that our plug-in design can be trained efficiently and boost the inference performance of various state-of-the-art acceleration methods, especially when there are few denoising steps. For example, when using 10 denoising steps on LSUN Bedroom dataset, we improve the FID of DDIM from 9.65 to 6.07, simply by adopting our method for a more appropriate set of timesteps. Code is available at \href{https://github.com/THU-LYJ-Lab/time-tuner}{https://github.com/THU-LYJ-Lab/time-tuner}.

cs.CV↗

arXiv2025

Orochi: Versatile Biomedical Image Processor

Deep learning has emerged as a pivotal tool for accelerating research in the life sciences, with the low-level processing of biomedical images (e.g., registration, fusion, restoration, super-resolution) being one of its most critical applications. Platforms such as ImageJ (Fiji) and napari have enabled the development of customized plugins for various models. However, these plugins are typically based on models that are limited to specific tasks and datasets, making them less practical for biologists. To address this challenge, we introduce Orochi, the first application-oriented, efficient, and versatile image processor designed to overcome these limitations. Orochi is pre-trained on patches/volumes extracted from the raw data of over 100 publicly available studies using our Random Multi-scale Sampling strategy. We further propose Task-related Joint-embedding Pre-Training (TJP), which employs biomedical task-related degradation for self-supervision rather than relying on Masked Image Modelling (MIM), which performs poorly in downstream tasks such as registration. To ensure computational efficiency, we leverage Mamba's linear computational complexity and construct Multi-head Hierarchy Mamba. Additionally, we provide a three-tier fine-tuning framework (Full, Normal, and Light) and demonstrate that Orochi achieves comparable or superior performance to current state-of-the-art specialist models, even with lightweight parameter-efficient options. We hope that our study contributes to the development of an all-in-one workflow, thereby relieving biologists from the overwhelming task of selecting among numerous models.

cs.CE↗

arXiv2025

Affine Frequency Division Multiplexing for Communication and Channel Sounding: Requirements, Challenges, and Key Technologies

Channel models are crucial for theoretical analysis, performance evaluation, and deployment of wireless communication systems. Traditional channel sounding systems are insufficient for handling the dynamic changes of channels in the next-generation space-air-ground-sea integrated networks (SAGSIN), which often results in outdated channel models that fail to provide reliable prior information for communication systems. To address this challenge, this paper proposes an integrated channel sounding and communication (ICSC) method as a practical solution. Unlike orthogonal frequency division multiplexing, affine frequency division multiplexing (AFDM) provides a full delay-Doppler representation of the channel, achieving optimal diversity in time-frequency doubly dispersive channels and effectively addressing the aforementioned challenges. Thus, we investigate the fundamental principles of AFDM, showing how it enables simultaneous communication and channel sounding, and explore key performance metrics for both functionalities. We also clarify the distinction and relationship between channel sounding, estimation, tracking and scatterer sensing. Additionally, several potential application scenarios for AFDM-ICSC are explored. Finally, we highlight the key challenges in implementing AFDM-ICSC, outline future research directions, and provide valuable insights for the continued development of this technology.

eess.SP↗

arXiv2025

CROP: Contextual Region-Oriented Visual Token Pruning

Current VLM-based VQA methods often process entire images, leading to excessive visual tokens that include redundant information irrelevant to the posed question. This abundance of unnecessary image details creates numerous visual tokens, drastically increasing memory and computational requirements in VLMs. To address this, we propose Contextual Region-Oriented Visual Token Pruning (CROP), a novel framework to compress visual tokens through a two-step process: Localization and Pruning. Specifically, CROP first employs an efficient model to identify the contextual region relevant to the input query. Subsequently, two distinct strategies are introduced for pruning: (1) Pre-LLM Compression (PLC), which adaptively compresses different image regions with varying ratios, and (2) Inner-LLM Pruning (ILP), a training-free method that prunes tokens within early LLM layers guided by the identified contextual region. Extensive experiments on a wide range of VQA tasks demonstrate that CROP significantly outperforms existing visual token pruning methods and achieves state-of-the-art performance.

cs.CV↗

arXiv2025

GTS_Forecaster: a novel deep learning based geodetic time series forecasting toolbox with python

Geodetic time series -- such as Global Navigation Satellite System (GNSS) positions, satellite altimetry-derived sea surface height (SSH), and tide gauge (TG) records -- is essential for monitoring surface deformation and sea level change. Accurate forecasts of these variables can enhance early warning systems and support hazard mitigation for earthquakes, landslides, coastal storm surge, and long-term sea level. However, the nonlinear, non-stationary, and incomplete nature of such variables presents significant challenges for classic models, which often fail to capture long-term dependencies and complex spatiotemporal dynamics. We introduce GTS Forecaster, an open-source Python package for geodetic time series forecasting. It integrates advanced deep learning models -- including kernel attention networks (KAN), graph neural network-based gated recurrent units (GNNGRU), and time-aware graph neural networks (TimeGNN) -- to effectively model nonlinear spatial-temporal patterns. The package also provides robust preprocessing tools, including outlier detection and a reinforcement learning-based gap-filling algorithm, the Kalman-TransFusion Interpolation Framework (KTIF). GTS Forecaster currently supports forecasting, visualization, and evaluation of GNSS, SSH, and TG datasets, and is adaptable to general time series applications. By combining cutting-edge models with an accessible interface, it facilitates the application of deep learning in geodetic forecasting tasks.

cs.LG↗