Searcharxiv⌕ Search

arXiv subjects

Tao Han

Publications and source records attributed to Tao Han.

At least 19 recordsLinked to original sources
arXiv2026

Temporally Grounded Compositional Camera Motion Understanding via Geometric Knowledge Distillation

Understanding camera motion is fundamental to video perception, with applications in spatial intelligence and controllable video generation. Multimodal large language models (MLLMs) provide a natural interface for this task, but existing work typically assigns one or more labels to an entire clip. Such clip-level recognition overlooks two defining properties of real camera motion: it can change within a shot, and multiple movements can occur simultaneously. We therefore formulate camera-motion understanding as temporally grounded, compositional recognition, which requires a model to localize motion-consistent intervals and identify every movement active within each interval. We introduce CamChoreo, a benchmark of 4,229 real single-shot clips with expert-annotated temporal segments. Its annotations use a compact vocabulary of 20 direction-aware labels, and nearly half of the segments contain compound camera motion, with multiple movement primitives active simultaneously. Recognizing such fine-grained, compositional motion is hard for current MLLMs, whose visual encoders emphasize semantic content rather than the geometric evidence on which camera motion depends. Directly injecting features from a frozen 3D foundation model addresses this gap, but requires running the expensive geometry model on every input; we refer to this baseline as CamInject. We instead propose CamDistill, which distills the same geometric knowledge into lightweight camera tokens during training and removes the 3D model at inference. CamDistill matches the accuracy of direct feature injection without running the 3D teacher at inference. Together, CamChoreo and CamDistill advance camera-motion understanding from clip-level labeling to temporally grounded, compositional recognition. Project page: https://ddz16.github.io/cammotion.github.io/.

cs.CV↗
arXiv2026

Archimedean Seesaw: Small Neutrino Masses and Large Lepton-number Violation

Contrary to the common lore that observable lepton-number violation (LNV) is inevitably suppressed by tiny neutrino masses, we identify a class of seesaw models in which arbitrarily large LNV can naturally coexist with sub-eV neutrino masses. We construct a symmetry-protected texture-zero structure in the neutrino Yukawa couplings and heavy Majorana mass matrix that gives rise to the required accidental symmetry, thereby protecting the light neutrinos from acquiring mass even in the presence of arbitrarily large LNV in the heavy sector. Small neutrino masses arise naturally from lifting the texture-zero structure while preserving the underlying symmetry. The resulting framework offers a rich and experimentally accessible phenomenology, predicting Heavy Neutral Leptons with sizeable active-sterile mixing over a broad range of experimentally accessible masses, giving rise to observable LNV signatures at collider and intensity-frontier experiments.

hep-ph↗
arXiv2026

Neutrino Fluxes at a Muon Collider

While muon colliders are primarily considered precision machines for new physics searches at the energy frontier, they are also intense sources of high-energy neutrinos. Large fluxes of electron and muon neutrinos are produced in the decay of beam muons, especially along the straight sections around the experiments. In this work, we re-evaluate these neutrino fluxes using the latest design of a \SI{10}{TeV} muon collider and find that fluxes around the interaction region are almost two orders of magnitude higher than earlier estimates. We also compute the fluxes of neutrinos produced in $\mu^+\mu^-$ collisions, electromagnetic showers induced by electrons from muon decay, and interactions of neutrinos close to the collider ring. We find that these additional neutrino sources are non-negligible and would lead to a sizable number of neutrino interactions of all flavors, including several tau neutrino events, in a ton-scale detector placed in the forward direction. We discuss the physics opportunities offered by muon collider neutrinos in the context of QCD and nuclear physics, electroweak precision measurements, and searches for new physics.

hep-ph↗
arXiv2026

Information-Theoretic Adaptive Cooling for Deterministic MPPI via Entropy Feedback

This paper investigates deterministic optimal control using Model Predictive Path Integral (MPPI) control, a sampling-based and derivative-free framework well suited for systems with complex dynamics and nonsmooth objectives. In deterministic MPPI, the temperature must be driven to zero to recover the true optimum, yet the design of an effective cooling schedule remains a fundamental challenge. Existing methods typically rely on predefined open-loop schedules, which limit the efficiency and robustness of the algorithm. To overcome this limitation, we propose an Information-Theoretic Adaptive Cooling (ITAC) framework that uses the Shannon entropy of the importance weights as an online feedback signal to regulate the temperature. The proposed mechanism adapts the cooling rate to the current sampling state, enabling fast progress when the weights are diffuse and cautious cooling when they become concentrated. We prove asymptotic convergence of the resulting scheme to the deterministic optimum, and further derive a critical entropy threshold that leads to a smooth barrier against premature weight collapse. Experiments on nonsmooth signal temporal logic motion-planning tasks show that ITAC improves sampling efficiency and achieves substantially faster convergence than state-of-the-art baselines without sacrificing the derivative-free nature of MPPI.

eess.SY↗
arXiv2026

TSSM: Triaxial State Space Model for Global Station Weather Forecasting with Temporal-Variable-Historical Modeling

Global Station Weather Forecasting (GSWF) is pivotal for localized and extreme weather prediction over key regions. Despite efforts to exploit look-back windows, existing methods show limited accuracy gains and struggle with extreme events and error accumulation. These limitations stem from overreliance on short-term patterns, which are insufficient to capture chaotic weather dynamics, especially under partial observations. To address this problem, we propose a novel Triaxial State Space Model (TSSM) with a history-enhanced Temporal-VariableHistorical paradigm, which incorporates period-aligned historical weather data to compensate for long-term, large-scale periodic, and full-window weather patterns beyond the temporal lookback window. Specifically, TSSM stacks historical samples into period-aligned batches, where forecasting is causally supported by historical and current observations. Temporal, variable, and historical scanning are designed to capture axial temporal dependencies, variable correlations, and historical evolution. This structure is hierarchically shared to model seasonal to extreme events while alleviating misalignment across historical patterns. TSSM achieves SOTA performance on Weather-5K, the largest station weather dataset to date, with 10% and 61% gains in accuracy and extreme event metrics, and obtains 95% best or second-best results on human-involved datasets. Its advantages are more pronounced in long-horizon and iterative forecasting, reaching a 37.5% gain at 240h and up to 103.5% under a 48h times 5 iterative setting. Moreover, TSSM retains > 90% performance under up to 80% missing observations, compared with < 43% for baselines, demonstrating robustness and practical potential for reliable GSWF in global in-situ observation networks.

cs.LG↗
arXiv2026

Decoherence and More Coherence in the Radiative Decay of the $Z$ Boson

Final state radiation in collider processes can be interpreted as interactions with unobserved degrees of freedom and is often discussed within the context of decoherence of an entangled state. We consider the radiative decay of the $Z$ boson as a representative example and perform a detailed analytical study of the $\tau^-\tau^+\gamma$ spin state, as determined by the chiral interactions of the Standard Model, over the complete three-body phase space. We explore various quantum information observables to quantify how the $\tau^-\tau^+$ spin state changes with the photon radiation. We find some striking features, for example that the emitted photon could either lead to decoherence or monotonic enhancement of entanglement for the fermion pair.

hep-ph↗
arXiv2026

RED-SEGA:Resilient Decentralized Stochastic Proximal Optimization with Gradient Sketching over Time-Varying Networks

Variance reduction is indispensable in Byzantine-resilient decentralized stochastic optimization over multi-agent systems (MASs) for its ability to mitigate gradient noise and thereby enhance the resilient aggregation process. However, most existing Byzantine-resilient decentralized variance-reduced (VR) stochastic gradient algorithms rely on random data sampling, which proves inefficient in data-scarce yet high-dimensional tasks, for instance, image deblurring. This paper pursues an alternative technical line that achieves variance reduction via gradient sketching. To this end, we first formulate a class of structural risk minimization (SRM) problems, where the local objectives are not necessarily decomposable and their gradients may be unavailable. To solve the SRM problems in a decentralized manner, we integrate a gradient-sketching technique into decentralized stochastic proximal gradient descent with gossip communication to propose a decentralized VR stochastic gradient algorithm, dubbed Gossip-SEGA.Since Gossip-SEGA does not provide any resilience against Byzantine attacks, a resilient extension of Gossip-SEGA,namely RED-SEGA,is developed via replacing the weighted average in Gossip-SEGA by a norm-penalized approximation. Theoretically, we derive sufficient conditions for both consensus (among reliable agents) and linear convergence rate of RED-SEGA over time-varying networks. The effectiveness and resilience of the proposed algorithms are validated through numerical experiments.

math.OC↗
arXiv2026

Quantum Information of Photon Pairs at Lepton Colliders

Photon pairs have provided an ideal laboratory for exploring entanglement and Bell inequality violation in low-energy experiments. Extending such studies to high-energy colliders is of great interest but has yet to be explored. Exploiting the photon conversion process for nearly on-shell photons, we formulate a factorization framework and an effective two-qubit description, which enable access to quantum information encoded in photon pairs. Using the existing Belle data set, we estimate that a $7.4\sigma$ violation of the Bell inequality could be achieved. The same framework can also probe quantum discord and nonstabilizerness, which could be measured with precisions of 5.6\% and 1.6\%, respectively. All the reconstructed results from photon conversion in the two-qubit framework are found to be consistent with the kinematic approach of real photons, and the formalism can apply to other spin-1 systems in an appropriate two-qubit limit.

hep-ph↗
arXiv2026

Robust Fall Recovery for Armless Bipedal-Wheeled Robots Via Force-Guided Learning

Fall recovery is critical for autonomous legged locomotion. Existing methods have demonstrated that some legged robots, such as humanoids and quadrupeds, are capable of fall recovery from diverse postures by utilizing arms or coordinating multi-legs to generate support forces. Without arms or other legs to provide supportive assistance, a bipedal-wheeled robot must rely solely on the actuation of its legs, making recovery particularly difficult. To address this, we introduce FTSR (Force-guided Teacher-student framework with Stage-wise Rewards). The force-guided method constructs an external auxiliary force during simulation training that correlates directly with the robot's real-time height, explicitly formulating this force as an optimizable constraint. Through constrained reinforcement learning, the policy is guided toward reducing force dependency gradually and increasing the body height, developing internal recovery strategies despite having no arms for support. Height-progressive stage-Wise rewards progressively structure posture stabilization during recovery and transition to sustained locomotion, integrated with teacher-student architecture distilling privileged knowledge of force effects and recovery dynamics. After simulation training, the policy is deployed on a physical armless bipedal-wheeled robot and extensively evaluated. Experiments confirm robust and reliable fall recovery under diverse challenging conditions, demonstrating strong environmental adaptability and motion robustness, while maintaining full post-recovery motion capability. The framework also generalizes effectively to a high-DOF humanoid, confirming its practical generalizability. The project page is available at https://2350575870.github.io/force-guided.github.io/

cs.RO↗
arXiv2026

AdaTok: Self-Budgeting Image Tokenization with Quality-Preserving Dynamic Tokens

Image tokenizers, from 2D grids to recent 1D sequences, typically encode every image with the same fixed number of tokens. Yet visual complexity is highly heterogeneous, so a uniform budget overspends on simple inputs and underserves complex ones. Existing elastic tokenizers expose variable-length reconstructions, but often leave token length as a deployment-time operating point, a search target, or an external prediction rather than an output of the tokenizer itself. In this work, we ask whether a discrete visual tokenizer can budget itself in one pass. Our central finding is that actionable elasticity requires a representation--allocation co-design: prefixes must remain decodable across budgets, and the tokenizer must learn which prefix each image needs. We propose AdaTok, a self-budgeting discrete 1D tokenizer. AdaTok combines Prioritized Representation Learning, which orders tokens with nested tail masking and resolves budget-dependent semantic shift through Multi-Head LoRA decoder heads, with Adaptive Token Allocation, which trains a lightweight deterministic-group GRPO policy over candidate budgets. Dynamic Pareto Weighting balances fidelity and efficiency during policy training without manual trade-off sweeps. On ImageNet-1K, AdaTok-Full reaches rFID 1.31 at 256 tokens, while AdaTok-Adaptive attains rFID 1.50 using only ~118 tokens on average, outperforming discrete 1D baselines at comparable budgets. In autoregressive image generation, the shorter adaptive representation yields ~2.1x throughput over a fixed 256-token decode, suggesting that visual token count can be learned as a content-conditioned output rather than set as a fixed hyperparameter.

cs.CV↗
arXiv2026

polyDAG: Polynomial Acyclicity Constraints for Efficient Continuous Causal Discovery in Visual Semantic Graphs

Modern image-analysis pipelines often convert images into structured semantic variables, such as facial attributes, object concepts, and scene descriptors. Learning directed dependencies among these variables can produce interpretable visual semantic graphs, but continuous directed acyclic graph learning is limited by the cost of enforcing acyclicity. We present polyDAG, a polynomial acyclicity framework for efficient continuous causal discovery in visual semantic graphs. polyDAG replaces the matrix-exponential acyclicity constraint with a finite polynomial trace constraint and proves that the new constraint is zero exactly for acyclic graphs. We further derive a geometric-series implementation that avoids the explicit summation loop while preserving the same acyclicity condition. Experiments on synthetic Erdos-Renyi graphs and CelebA facial visual attributes show that polyDAG improves efficiency and structure recovery. Averaged over the revised synthetic protocol with d in {100, 200, 500}, polyDAG reduces mean structural Hamming distance from 318.4 to 285.4 and improves mean F1 score from 0.725 to 0.756. At 100 nodes, the geometric variant runs in 3.44 seconds compared with 5.16 seconds for the exponential baseline, corresponding to a 33.4 percent speedup. Code and data are publicly available at https://github.com/wenhaoz-fengcai/polyDAG.

cs.CV↗
arXiv2026

Graph Attention-Based Virtual Metrology for Film Deposition Processes in Semiconductor Manufacturing

Artificial intelligence-driven semiconductor manufacturing increasingly operates at nanometer and angstrom scales, where precise process control depends on accurate and timely metrology. However, physical metrology is limited by measurement latency, cost, and sampling constraints, restricting its scalability in high-volume production. Virtual metrology (VM) has emerged as an effective alternative by predicting wafer-level characteristics from equipment sensor data. Despite recent advances, many existing VM models remain correlation-driven and lack the ability to capture structured dependencies among heterogeneous process variables, while providing limited interpretability. This study presents a graph attention-based VM framework for film deposition processes that integrates temporal feature learning with structured parameter-layer dependency modeling. The proposed approach represents each step-parameter pair as a node and extracts temporal embeddings from high-frequency equipment traces using convolutional feature encoders. A parameter-to-layer graph attention mechanism is employed to model directional dependencies, enabling each film layer to aggregate relevant process information. The framework is evaluated using industrial deposition data collected from production wafers, where the model predicts film thickness from multivariate sensor signals. Experimental results demonstrate improved predictive performance compared to baseline models. In addition, analysis of the learned attention weights reveals interpretable parameter-layer relationships consistent with physical process behavior, capturing dominant process factors and temporal dependencies across deposition stages. These results indicate that the proposed framework enhances prediction accuracy and provides meaningful insight into process dynamics, supporting effective monitoring and optimization in semiconductor manufacturing.

cs.CE↗
arXiv2026

Train the Agent, Not the Expert: Learning to Harness Heterogeneous Experts for Multi-Turn Visual Reasoning

Recent progress in computer vision has produced a wide range of powerful specialized models for detection, segmentation, counting, and other visual tasks. However, these models are usually optimized for isolated task formulations, making it difficult to directly support general-purpose visual intelligence, especially when a task requires complex language understanding and dense small-object perception. In this paper, we propose VisHarness, a trainable visual agent that decouples high-level perception, reasoning, and decision-making from low-level task execution. Instead of training a model to solve a specific visual task, VisHarness learns to harness a set of carefully designed heterogeneous visual experts. This paradigm preserves the general intelligence of the agent while fully leveraging the precision advantages of specialized visual models in concrete visual tasks. With only lightweight training, VisHarness learns a generalizable visual expert-harnessing policy and can solve common fundamental vision tasks under various complex conditions through multi-turn interactions with visual expert models. To enable efficient on-policy reinforcement learning training in a live environment, we introduce dynamic visual memory archiving, which mitigates the rapidly accumulating visual-token overhead caused by multi-turn interactions with visual expert models. Experiments on four representative benchmarks covering reasoning segmentation, generalized referring segmentation, dense small-object detection, and referring counting demonstrate that VisHarness substantially outperforms existing general-purpose models and achieves competitive or superior performance compared with task-specific models.

cs.CV↗
arXiv2026

RealBench: Benchmarking Data-Driven Numerical Weather Forecasting Under Operational Conditions and Extreme Event Challenges

Accurate evaluation of weather forecasting models is critical for their reliable deployment in real-world applications. However, existing benchmarks predominantly rely on reanalysis products such as ERA5, which are generated through delayed data assimilation and do not reflect the constraints of real-time operational forecasting, thereby resulting in a systematic mismatch between benchmark performance and real-world forecasting. In this work, we introduce RealBench, a next-generation benchmark for AI weather forecasting that emphasizes realistic evaluation under operational conditions. RealBench features a strictly out-of-distribution test set spanning 2025 to eliminate data leakage and capture recent atmospheric regimes. It integrates multiple data sources, including low-latency operational analysis and a large-scale global in-situ observation dataset comprising over 10,000 stations, enabling direct evaluation against real atmospheric measurements. Beyond standard global metrics, RealBench provides a comprehensive evaluation framework for high-impact extreme events, including heatwaves, cold surges, and tropical cyclones, using event-specific metrics that better reflect real-world forecasting priorities. The evaluation results reveal substantial discrepancies between reanalysis-based metrics and real-world performance, particularly concerning extreme events. By highlighting the limitations of existing benchmarks, this work establishes a more faithful and operationally relevant evaluation paradigm, providing a rigorous foundation for advancing next-generation AI weather forecasting systems. The benchmark implementation is available at: https://github.com/lixruize-del/NWP-Benchmark.

cs.LG↗
arXiv2026

MLLMs Know When Before Speaking: Revealing and Recovering Temporal Grounding via Attention Cues

Video temporal grounding (VTG), which localizes the start and end times of a queried event in an untrimmed video, is a key test of whether multimodal large language models (MLLMs) understand not only what happens but also when it happens. Although modern MLLMs describe video content fluently, their timestamp predictions remain unreliable, while existing remedies either require costly post-training on temporal annotations or rely on coarse training-free heuristics. In this work, we probe the cross-modal attention of MLLMs and uncover a perception-generation gap. Our key finding is that MLLMs often know the target interval during prefill, but lose this signal when generating the final answer. In the prefill stage, a sparse set of attention heads, which we call \emph{Temporal Grounding Heads} (TG-Heads), concentrates query-to-video attention on the ground-truth interval. During autoregressive decoding, however, the answer tokens shift attention away from this interval toward visually salient but query-irrelevant segments. This observation motivates an inference-time read-then-regenerate framework. We first convert TG-Head prefill attention into a debiased frame-level relevance signal and extract the high-attention interval it highlights. We then re-invoke the MLLM with visual context restricted to this interval, using video cropping or attention masking to suppress distractors. Without parameter updates and architectural changes, our framework consistently improves MiMo-VL-7B, Qwen3-VL-8B, and TimeLens-8B on three VTG benchmarks, with gains of up to +3.5 mIoU. The project website can be found at https://ddz16.github.io/mllmsknowwhen.github.io/.

cs.CV↗
arXiv2026

Learning Spatiotemporal Sensitivity in Video LLMs via Counterfactual Reinforcement Learning

Video large language models (Video LLMs) achieve strong benchmark accuracy, yet often answer video questions through shortcuts such as single-frame cues and language priors rather than by tracking spatiotemporal dynamics. This issue is exacerbated in RL post-training, where correctness-only rewards can further reinforce shortcut policies that obtain high reward without tracking video dynamics. We address this by asking a controlled counterfactual question: if the visual world changed while the question remained fixed, should the answer change or stay the same? Based on this view, we propose \textbf{Counterfactual Relational Policy Optimization (CRPO)}, a dual-branch RL framework for improving \emph{spatiotemporal sensitivity}. CRPO constructs counterfactual videos through horizontal flips and temporal reversals, trains on both original and counterfactual branches, and introduces a \textbf{Counterfactual Relation Reward (CRR)} between their answers. CRR encourages answers to change for dynamic questions and remain unchanged for static questions. This cross-branch constraint makes it difficult for shortcut policies to be consistently rewarded across both branches. To evaluate this property, we introduce \textbf{DyBench}, a paired counterfactual video benchmark with 3,014 videos covering reversible dynamics, moving direction, and event sequence, together with a strict pair-accuracy metric that prevents fixed-answer shortcuts from inflating scores. Experiments show that CRPO outperforms prior RL methods on spatiotemporal-sensitive evaluations while maintaining competitive general video performance. On Qwen3-VL-8B, CRPO improves DyBench P-Acc by +7.7 and TimeBlind I-Acc by +8.2 over the base model, indicating improved spatiotemporal sensitivity rather than stronger reliance on static shortcuts. The project website can be found at https://ddz16.github.io/crpo.github.io/ .

cs.CV↗
arXiv2026

EdgeServing: Deadline-Aware Multi-DNN Serving at the Edge

As edge computing expands, serving multiple deep neural network (DNN) models on a single shared GPU has become a common yet challenging scenario, where each scheduling decision affects the tail latency of all concurrent queues. Existing schedulers rely on local heuristics and fail to capture this global impact, while GPU spatial-sharing approaches sacrifice latency predictability. In this paper, we propose EdgeServing, a deadline-aware multi-DNN serving system for edge devices. EdgeServing adopts time-division GPU sharing with early-exit inference for high inference predictability, and introduces a stability score to quantify how each candidate scheduling decision impacts the future queue status. At runtime, it cohesively selects the model, exit point, and batch size to minimize predicted system-wide SLO impact. Experimental results on multiple hardware platforms show that EdgeServing consistently outperforms representative baselines in both SLO violation ratio and P95 latency, enabled by early-exit mechanism, which expands the scheduling action space under tight latency constraints.

cs.DC↗
arXiv2026

Earth-o1: A Grid-free Observation-native Atmospheric World Model

Despite the unprecedented volume of multimodal data provided by modern Earth observation systems, our ability to model atmospheric dynamics remains constrained. Traditional modeling frameworks force heterogeneous measurements into predefined spatial grids, inherently limiting the full exploitation of raw sensor data and creating severe computational bottlenecks. Here we present Earth-o1, an observation-native atmospheric world model that overcomes these structural limitations. Rather than relying on conventional atmospheric dynamical modeling systems or traditional data assimilation, Earth-o1 directly learns the continuous, three-dimensional physical evolution of the Earth system from ungridded observational data. By integrating diverse sensor inputs into a unified, grid-free dynamical field, the model autonomously advances the atmospheric state in space and time. We show that this fundamentally distinct paradigm enables direct, real-time forecasting and cross-sensor inference without the overhead of explicit numerical solvers. In hindcast evaluations, Earth-o1 achieves surface forecast skill comparable to the operational Integrated Forecasting System (IFS). These results establish that continuous, observation-driven world models -- a new class of fully observation-native geophysical simulators -- can match the fidelity of established physical frameworks, providing a scalable data-driven foundation for a digital twin of the Earth.

cs.CV↗