Searcharxiv⌕ Search

arXiv subjects

Lei Zhu

Publications and source records attributed to Lei Zhu.

At least 19 recordsLinked to original sources
arXiv2026

Rethinking Underwater Image Enhancement: Learning through Text-Guided Degradation Curriculum

Existing paired clean-degraded datasets have advanced learning-based underwater image enhancement, but they remain either limited in scale or difficult to acquire, and still fail to adequately cover the diverse degradations encountered in real-world underwater environments. Such scarce and insufficient data ultimately restrict the generalization of underwater image enhancement models, highlighting the need for realistic and controllable degradation modeling to better capture real underwater conditions and improve enhancement robustness. To this end, we propose a unified degradation-aware training framework for the underwater image enhancement task. A Text-Conditioned Degradation Augmentation Pipeline uses a text-guided diffusion model to synthesize diverse and semantically consistent degraded samples, providing paired data to support the subsequent curriculum learning strategy. The proposed Progressive Degradation Curriculum organizes training across multiple attributes and difficulty levels to promote balanced learning, and a Transmission-Guided One-Step Diffusion Model leverages this progressive process to better address the wide range of degradation severity and restoration ambiguity in underwater images. Extensive experiments on multiple underwater benchmarks demonstrate that our proposed method achieves superior performance. Our code and models will be released upon the publication of this work.

cs.CV↗
arXiv2026

The ALMA-QUARKS survey: A Keplerian Circumbinary Disk Candidate in the High-mass Star-forming Region IRAS 08303-4303

We present the ALMA Band 6 continuum and molecular line observations of IRAS 08303-4303, a high-mass star-forming region, with an angular resolution of $\sim 0.3''$ ($\sim 700$ au at the distance of 2.35 kpc). Six dense condensations are identified in the 1.3 mm continuum, and at least four of them are associated with molecular outflows. The estimated outflow parameters, particularly mass outflow rates $\geq 10^{-4}$ $M_{\odot}$ yr$^{-1}$, suggest that the driving sources are high-mass young stellar objects. Two of the condensations are forming a wide protobinary system with a separation of $\sim 900$ au. They reside within a single rotating structure, identified as a circumbinary disk which has a mass of about 2 $M_{\odot}$. This structure is traced by multiple molecular lines and exhibits a Keplerian rotation pattern. Based on the Keplerian rotation assumption, the enclosed mass is derived to be between 32 and 61 $M_{\odot}$: 22-43 $M_{\odot}$ for the primary and 10-18 $M_{\odot}$ for the secondary. The circumbinary disk is on the whole gravitationally stable, yet some parts with Toomre $Q < 2$ indicate potential fragmentation. Various molecular lines reveal a streamer-like structure connecting to the circumbinary disk, suggesting transfer of material from the envelope to the protobinary system through it. Despite the observational rarity of such circumbinary disks around high-mass YSOs, the high binary fraction among massive stars underscores their significance. This source, therefore, serves as a key laboratory for understanding the formation of high-mass binary systems.

astro-ph.SR↗
arXiv2026

The ALMA-QUARKS Survey: Discovery of Superfine Dusty Fibrils inside Massive Star-forming Clumps

We report the discovery of more than 320 superfine dusty filamentary structures (fibrils) inside 121 massive star forming clumps that are located in widely different Galactic environments (Galactocentric distances of ~1.2 - 12.9 kpc). These fibrils are identified from the 1.3-mm continuum emission in the ALMA-QUARKS survey, which has a linear resolution of ~900 AU for a source at ~3 kpc. The typical width of these fibrils identified in clumps is found to be ~0.01 pc, which is about ten times narrower than that of dusty filaments in nearby clouds identified by the Herschel Space Observatory. The mass (M) versus length (l) relation for these fibrils follows M ~ l^2, similar to that of Galactic filaments identified in space (e.g., Herschel) and ground-based single-dish (e.g., APEX) surveys. However, these fibrils in the ALMA-QUARKS survey are significantly denser (N_{H_2} = 10^{23} - 10^{24} cm^{-2}) than the filaments found in previous Herschel surveys (N_{H_2} = 10^{20} - 10^{23} cm^{-2}). This work contributes a large sample of newly found superfine fibrils in massive clumps, following the identification of large 0.1 pc-wide filaments and associated internal velocity coherent molecular fibers in nearby molecular clouds, further emphasizing the crucial role played by filamentary structures in star formation at various physical scales.

astro-ph.GA↗
arXiv2026

PixelWizard: Towards Efficient High-Fidelity Video Generation at Ultra-Large Spatial Resolution

High-resolution video generation faces a coupled bottleneck of optimization instability and prohibitive computational costs. The massive expansion of the token sequence not only biases optimization toward local textures at the expense of global coherence, leading to structural collapse, but also imposes prohibitive training costs and severe inference latency. To address this, we propose PixelWizard, a framework that hierarchically decouples global structure modeling from fine-grained detail synthesis. PixelWizard first establishes a compact spatiotemporal anchor to concentrate dense structural priors, which then guides fine-grained generation at high resolution. This mitigates the local optimization bias to ensure structural stability without compromising high-frequency details. Leveraging this structural stability, we introduce Noise-Span Aligned Shortcut Training to break the inference bottleneck. By explicitly modeling the step size, this mechanism allows the model to traverse the generation trajectory with large steps. Crucially, we incorporate Exponential Index-Biased Sampling and Adaptive Noise-Span Calibration to align optimization with the shifted noise schedules of high-resolution grids, ensuring robust few-step inference without incurring the heavy overhead of distillation. Extensive experiments demonstrate that PixelWizard achieves superior visual quality while accelerating the generative sampling of native 2K/4K videos by over 10x.

cs.CV↗
arXiv2026

UltraDiff: Transferring High-Fidelity Priors to Compressed Latent Spaces for High-Resolution Image Generation

High-resolution synthesis with Latent Diffusion Models (LDMs) faces a fundamental trade-off between computational cost and generative fidelity. This trade off is largely governed by the latent compression factor f (e.g. f8 denotes f=8 and f32 f=32). While aggressive compression (f32) significantly improves efficiency, it suffers from severe information loss and ambiguity compared to standard f8 models. We propose UltraDiff, a framework that reformulates this challenge as joint latent diffusion learning followed by marginal score distillation. We introduce MixTeacher, a dual branch DiT that models the joint f8-f32 distribution, utilizing shared attention to resolve the ambiguity of the compressed space with high fidelity f8 priors. Crucially, we propose vector field distillation in the native f32 space. Theoretically, this compels the student to learn the conditional expectation of the teacher trajectory, effectively filtering out compression artifacts. Across the evaluated 1K to 4K settings, UltraDiff approaches f8 level quality while preserving f32 efficiency. With 20 sampling steps on a single A100 GPU, it generates a 2K image in 6.2 seconds over 40 times faster than our f8 baseline.

cs.CV↗
arXiv2026

Counterfactual Neighborhood Risk Fields: Evidence-Grounded Selective Safety for Vision-Language Models

Representation-based safety defenses for vision-language models (VLMs) often learn to separate harmful from benign inputs, but separation alone does not establish what a risk score means or when it should be trusted. We introduce Counterfactual Neighborhood Risk Fields (CNRF), an evidence-grounded framework for selective VLM safety. CNRF constructs context-matched benign-harmful counterfactual pairs, uses their representation differences as capability-related risk directions, and estimates neighborhood support separately from risk. This separation prevents unfamiliarity from being conflated with harmfulness and enables selective intervention rather than blanket rejection. A victim-adaptive interface accommodates different safety observations across VLM architectures and routes selected requests for same-victim response review, which either releases acceptable candidates or produces constrained alternatives with reduced harmful capability. Across three VLMs and five multimodal jailbreak families, CNRF achieves 86.06-89.30% macro defense success rate while retaining substantial benign-task utility. Controlled experiments further show that query-matched references and preserved context-direction correspondence materially improve risk ranking in the low-false-positive regime. CNRF thus reframes representation-based VLM safety from learning harmfulness boundaries to constructing applicable evidence for selective risk decisions.

cs.AI↗
arXiv2026

The ALMA-QUARKS Survey. IV. Core Mass Function and Evolutionary Trends in Core Formation Efficiency and Chemistry

The ALMA-QUARKS (Querying Underlying Mechanisms of Massive Star Formation with ALMA-Resolved Gas Kinematics and Structures) survey targets 139 massive protoclusters selected from the ATOMS survey. The targets were observed at 1.3~mm wavelength with an angular resolution of $\sim0.3^{\prime\prime}$, corresponding to $\sim10^3$~au at the sample's median distance of $\sim3.8$~kpc. Using \texttt{getsf}, we identify 3374 dense cores and classify them as starless, protostellar, hot core, or H{\sc ii} region candidates based on spectral-line diagnostics (e.g., H30$α$, CH$_3$CN, and CH$_3$OH) and CO outflow associations. The resulting catalog provides core physical properties and enables statistical investigations of dense core populations in massive protocluster environments. The derived global core mass function (CMF) after excluding H{\sc ii} region candidates, which is biased by distance effects, has a relatively flat power-law slope at the high-mass end ($\mathrm{d}N/\mathrm{d}\log M \propto M^{-γ}$, with $γ=0.98\pm0.03$). In contrast, the measured high-mass slopes in individual distance bins are approximately Salpeter-like, while their fitted lower mass limits increase with distance. We further examine the core formation efficiency (CFE) across the full sample and find that it positively correlates with the luminosity-to-mass ratio of the parental clumps, consistent with an evolutionary increase in the fraction of mass concentrated into dense cores. We also investigate the detection rates of different molecular tracers and find systematic chemical variations with evolutionary stage. Compared with the ASHES survey, we find that the detection rates of deuterated species vary significantly even for cores at the same evolutionary stages, suggesting an important environmental influence on the chemical evolution of cores.

astro-ph.GA↗
arXiv2026

OmniVBench: A Benchmark and Large-Scale Dataset for Omni Reference-to-Video Generation

Omni reference-to-video (R2V) generation is emerging as a general paradigm that enables flexible video customization from diverse and compositional visual references. However, existing benchmarks fall short of these emerging capabilities: their test cases cover only a limited range of reference types and compositions, and their evaluation protocols largely assess holistic reference consistency, overlooking whether reference factors are properly preserved, disentangled, and routed. Meanwhile, the high cost of constructing omni R2V training data makes suitable training resources scarce. To address these gaps, we introduce OmniVBench and the Omni-R2V Dataset, establishing a shared foundation for evaluating and training omni R2V models. OmniVBench expands R2V evaluation across broader reference types, richer reference compositions, and more flexible instruction-driven customization, covering 7 task families and 18 fine-grained tasks. Complementing this task design, we introduce factor-grounded evaluation with 12,172 case-specific checklist items, explicitly assessing whether intended reference factors are faithfully preserved, correctly disentangled and bound to their targets, and realized according to the instruction. We further introduce the Omni-R2V Dataset, bringing industrial-grade training resources for diverse R2V tasks to the broader research community. Drawing primarily on a large-scale corpus of professional video footage, it comprises 340K processed training samples spanning diverse reference types and multi-reference compositions. Specifically, we develop task-specific pipelines for reference-target pair construction, offering a practical and scalable recipe for omni R2V data construction. Extensive evaluation of advanced open- and closed-source R2V models reveals substantial performance variation across diverse tasks and distinct capability patterns across evaluation dimensions.

cs.CV↗
arXiv2026

Tikhonov-Stabilized Bezier Representation Forecasting for Training-free Diffusion Acceleration

Diffusion models, particularly Diffusion Transformers, achieve strong image and video generation quality but remain expensive at inference time due to repeated denoiser evaluations. Feature caching offers a deployment-friendly acceleration strategy by reusing or predicting intermediate representations without retraining the generator or modifying the sampler. However, existing cache-then-forecast methods often rely on Taylor-style finite-difference extrapolation, which can become unstable over longer cache intervals, and typically predict local module outputs whose errors may accumulate through subsequent denoiser blocks. We propose \textbf{BeziCast}, a training-free diffusion acceleration framework that forecasts output-proximal denoising representations using low-order B'ezier trajectories. BeziCast estimates B'ezier control points via Tikhonov-stabilized fitting, providing a smooth, capacity-controlled temporal parameterization that avoids explicit high-order derivative estimation and decouples trajectory capacity from cache-query density. To moderate aggressive extrapolation, we further derive equivalent forecasting weights and introduce a convex-hull-inspired guardrail that detects high-risk predictions and softly pulls them toward a conservative simplex-projected estimate. Extensive evaluations across advanced image and video diffusion models demonstrate the effectiveness of BeziCast. In particular, BeziCast accelerates FLUX.1 by up to (4.79\times) and HunyuanVideo by up to (4.11\times), while preserving substantially better generation quality than competing acceleration baselines.

cs.CV↗
arXiv2026

Physics-Informed Conditional Diffusion for Motion-Robust Retinal Temporal Laser Speckle Contrast Imaging

Retinal laser speckle contrast imaging (LSCI) is a noninvasive optical modality for monitoring retinal blood flow dynamics. However, conventional temporal LSCI (tLSCI) reconstruction relies on sufficiently long speckle sequences to obtain stable temporal statistics, which makes it vulnerable to acquisition disturbances and limits effective temporal resolution. A physically informed reconstruction framework, termed RetinaDiff (Retinal Diffusion Model), is proposed for retinal tLSCI that is robust to motion and requires only a few frames. In RetinaDiff, registration based on phase correlation is first applied to stabilize the raw speckle sequence before contrast computation, reducing interframe misalignment so that fluctuations at each pixel primarily reflect true flow dynamics. From the long registered sequence this step yields a high-quality multiframe tLSCI map that serves only as the reconstruction target, while a motion-corrected contrast prior is computed independently from the few input frames. Next, guided by this prior, a conditional diffusion model performs inverse reconstruction by jointly conditioning on the registered few-frame sequence and the prior. On stable sequences acquired with an in-house retinal LSCI system, RetinaDiff improved SSIM from 0.159 to 0.533, PSNR from 14.83 to 18.06 dB, and FID from 211.50 to 111.55 compared with direct five-frame reconstruction, showing improved structural continuity and statistical stability over representative baselines. The framework also remains effective in a small number of extremely challenging cases, where both the direct five-frame input and the conventional multiframe reconstruction are severely degraded. Overall, this work provides a practical and physically grounded route for reliable retinal tLSCI reconstruction from extremely limited frames. The source code and model weights will be released upon acceptance.

cs.CV↗
arXiv2026

Two Margins of Climate Cooperation: Emissions and Coalition Membership under Tipping

We study climate coalition stability with endogenous temperature and stochastic tipping. In our dynamic game, rising temperature widens the emissions gap between members and nonmembers, shrinking stable coalitions over time. The threat of tipping strengthens free-riding, raising emissions and reducing coalition size; its occurrence also contracts coalitions numerically. Advanced regions form the core, while emerging & developing regions drive membership variation. Comparing policy instruments, connection (internal incentives) and sanctions (external penalties) sustain similar coalitions, but connection raises aggregate payoffs, whereas sanctions achieve larger emissions cuts. As tipping approaches, this trade-off intensifies, making the optimal institutional design state-dependent.

econ.GN↗
arXiv2026

SoL-Refiner: Speed-of-Light One-Step Refinement for High-Resolution Video

High-resolution video generation is expensive, as its cost grows rapidly with the number of spatiotemporal tokens. A practical alternative first generates a lower-resolution video and then applies a refiner, but conventional multi-step refinement introduces a second sampling bottleneck. We present SoL-Refiner, a one-step video refiner that transforms low-resolution model outputs into 4K videos with a single denoising step. Our three-stage recipe combines high-resolution continual training, reinforcement learning (RL) post-training, and a final one-step distillation. We introduce Refiner-Bench, a video refinement benchmark constructed from the outputs of different video generators, and use a shared-input protocol to compare refiners at approximately 2K output resolution. At 2K, the one-step SoL-Refiner outperforms all external refiners on the VBench and UniPercept averages, while at $3840\!\times\!2176$ it improves both metrics over the three-step LTX-2.3 Refiner. With the complete acceleration stack, SoL-Refiner achieves an $8.91\times$ speedup in refinement latency over the same baseline in our 2K latency setting.

cs.CV↗
arXiv2026

Low latency global carbon budget reveals strong land sink recovery in 2025

The atmospheric CO2 growth rate fell sharply in 2025, from a record 3.76 $\pm$ 0.09 ppm yr-1 in 2024 to 2.06 $\pm$ 0.09 ppm yr-1 (NOAA marine boundary layer observations), below the 2015-2022 mean of 2.47 ppm yr-1, even as fossil CO2 emissions rose by 0.7% to 10.38 GtC yr-1. Here we present a low-latency global and regional carbon budget for 2025, combining three dynamic global vegetation models (DGVMs) and ocean model emulators with four atmospheric inversions constrained by OCO-2 satellite retrievals. The global net land sink reached 2.36 $\pm$ 0.16 GtC yr-1 in 2025 (DGVMs: 2.04 $\pm$ 0.24; inversions: 2.68 $\pm$ 0.20 GtC yr-1), strengthening by 2.81 $\pm$ 0.31 GtC yr-1 from 2024 and exceeding the 2015-2022 mean by 0.71 $\pm$ 0.13 GtC yr-1. Ocean uptake (3.11 $\pm$ 0.36 GtC yr-1) remained similar to 2024, making the land sink rebound the dominant driver of the slowdown in CO2 growth. Tropical lands shifted from net sources in 2024 to net sinks in 2025, with enhanced uptake across much of Africa and northern Eurasia, and land flux anomalies covaried with GRACE terrestrial water storage. Where the sink had weakened substantially in 2023-2024, about 80% of the area showed some recovery, with overall recovery of 87.3% (DGVMs) to 99.5% (inversions). Recovery exceeded 100% in the tropics but remained incomplete in the northern extratropics, indicating a strong but spatially uneven rebound of the land carbon sink.

physics.ao-ph↗
arXiv2026

Robust small-molecule identification from incomplete, degraded, and inconsistent spectra using multimodal mixed-condition training

Reliable small-molecule identification often requires complementary evidence from multiple spectroscopic measurements. In practice, however, spectra may be unavailable, degraded by measurement-related variations, or even incorrectly associated with a sample, thereby hindering accurate molecular identification. Herein, we propose a multimodal mixed-condition training strategy that accommodates missing, degraded, and mismatched measurements for small-molecule structure identification. The strategy incorporates chemical and spectroscopic knowledge through predefined missing-input configurations, modality-specific spectral perturbations, and chemically informed spectrum replacements. Models were trained on 635,441 samples comprising mass spectrometry (MS), infrared (IR), and nuclear magnetic resonance (NMR) simulated spectra from the Multimodal Spectroscopic Dataset (MSSD). They were then systematically evaluated on 79,462 held-out samples across 30 views designed to represent variations in spectra. A controlled comparison of complete-input and mixed-condition training under concatenation and mixture-of-experts (MoE) fusion showed that the training strategy was the principal source of improvement. For MoE, mixed-condition training increased the mean reciprocal rank (MRR) by 6.08% (from 0.9203 to 0.9763) and the top-1 molecular identification rate by 7.67% (from 89.50% to 96.36%). Notably, under single-modality inputs, IR MRR increased 2.15-fold (from 0.4337 to 0.9307), while MS MRR increased 2.31-fold (from 0.3711 to 0.8575). With the proposed strategy, complete-input performance remained high, while sample-level mismatch detection also improved. Together, these results highlight the potential of multimodal mixed-condition training for practical molecular identification by explicitly addressing incomplete, degraded, and mismatched measurements encountered in real-world analysis.

cs.LG↗
arXiv2026

SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness

As coding agents move from supervised code completion to unattended, around-the-clock exploration, their work expands from isolated predictions into long trajectories of reasoning, tool use, and feedback. Token efficiency therefore becomes important for scaling recursive self-improvement. We take an RSI-inspired approach at the harness layer, scaling auto-research loops across increasingly numerous and diverse environments for harness rollouts. At this scale, the process yields reusable improvements that transfer beyond their development setting, moving automated harness discovery toward production-level outcomes. Four mechanisms survive selection and form SoL-Pi, spanning action execution, context compaction, observation handling, and delegated reading. On the 51-task EdgeBench evaluation, SoL-Pi achieves performance comparable to Pi across GPT-5.6 Sol and Opus 5 while reducing recorded token traffic by 44.7-49.0% and API cost by about one third. In other words, estimated hourly savings are \$8.75-\$13.50 relative to native Codex and Claude Code harnesses, and \$4.36-\$5.71 relative to Pi.

cs.AI↗
arXiv2026

Towards Generalization of Block Attention via Automatic Segmentation and Block Distillation

Block attention, which processes the input as separate blocks that cannot attend to one another, offers significant potential to improve KV cache reuse in long-context scenarios such as Retrieval-Augmented Generation (RAG). However, its broader application is hindered by two key challenges: the difficulty of segmenting input text into meaningful, self-contained blocks, and the inefficiency of existing block fine-tuning methods that risk degrading performance. To address these, we first construct SemanticSeg, a large and diverse semantic segmentation dataset containing over 30k instances across 16 categories-including books, code, web text, and conversations with text lengths ranging from 2k to 32k. Using this dataset, we train a lightweight segmenter to automatically partition text into human-instinct-aligned blocks with controllable granularity. Second, we propose block distillation, a training framework that is more efficient than block fine-tuning, which uses a frozen full-attention teacher model to guide the block-attention student. This framework integrates three novel components: block sink tokens to mitigate information loss at block boundaries, block dropout to leverage training signals from all blocks, and token-level loss weighting to focus learning on block-attention-sensitive tokens. Experiments across multiple models and benchmarks demonstrate that our segmenter outperforms heuristic and statistical baselines, and block distillation achieves near-full-attention performance under block attention, establishing a practical and scalable pathway for deploying block attention.

cs.CL↗
arXiv2026

SAS: Simple Attention Sparsification via End-to-End Optimization of Context Ranking

Post-training attention sparsification reduces the quadratic cumulative attention cost of pretrained Transformers by selecting a small set of context units (tokens or blocks) for each query. Existing trainable methods usually use a lightweight selector to score context units, followed by hard Top-K selection that blocks gradients from the language modeling loss. Consequently, these methods commonly distill layer-wise dense attention distributions. Although this encourages the selector to rank context units by dense attention weights in the original model, the ranking is not directly aligned with their impact on predictions under a fixed attention budget (i.e., the number of attended context units per query), potentially wasting the limited budget on less useful units. To address this misalignment, we propose Simple Attention Sparsification (SAS), a gated sparse attention mechanism that optimizes context ranking end-to-end with the language modeling loss. The key idea is to inject the selector's continuous scores into attention logits during training, allowing the loss to update the selector through standard backpropagation. We identify several choices crucial for this simple design to work well in practice: placing the gate inside the attention softmax in log form, using normalized softmax gates to calibrate historical context against the always-retained current block, and preserving continuous selector scores so the model learns relative priorities rather than only hard selections. To support long-sequence training, we implement a memory-efficient Triton kernel that integrates SAS into FlashAttention-style computation. Across reasoning, long-context understanding, and agentic tasks, SAS consistently outperforms trainable sparse attention baselines across attention budgets, with especially large gains under tight budgets, demonstrating more effective context ranking for downstream tasks.

cs.CL↗
arXiv2026

ViT3Flow: A Test-Time Training Transformer MeanFlow for Postoperative Radiograph Synthesis in Scoliosis

Predicting postoperative spinal morphology from preoperative radiographs could provide valuable support for scoliosis surgical planning, but remains challenging because surgical correction induces large spatial changes while anatomical structures must be faithfully retained. We formulate this problem as postoperative scoliosis radiograph synthesis and construct ScoliSurg, the first paired dataset for this task, comprising 632 preoperative--postoperative whole-spine radiograph pairs with structured morphology information. We further propose ViT$^{3}$Flow, a single-NFE conditional MeanFlow framework for efficient postoperative radiograph synthesis. ViT$^{3}$Flow models surgical correction as finite-interval generative transport and replaces conventional self-attention with test-time-training token mixers that perform sample-specific inner adaptation to the anatomy and deformity pattern of each case. In addition, a Spinal Morphology Extraction Agent extracts distributions of dominant-curve region and direction from the preoperative radiograph. These distributions guide Diagnosis-Routed Interval Cross-Attention (DRICA), which performs interval-dependent vertical, horizontal, joint, and global retrieval from a separate preoperative token stream. This design enables the evolving postoperative representation to incorporate spatially corresponding anatomical evidence throughout the transport process. Extensive experiments on ScoliSurg demonstrate that ViT$^{3}$Flow achieves the best performance among the compared methods in perceptual image quality, anatomical fidelity, and clinically relevant geometric accuracy, while requiring only a single network evaluation. These results highlight the potential of ViT$^{3}$Flow for efficient and anatomically faithful postoperative radiograph synthesis in scoliosis surgical planning.

cs.CV↗