arXiv subjects

Yi Liu

Publications and source records attributed to Yi Liu.

At least 325 records · Page 18Linked to original sources

arXiv2025

PWD: Prior-Guided and Wavelet-Enhanced Diffusion Model for Limited-Angle CT

Generative diffusion models have received increasing attention in medical imaging, particularly in limited-angle computed tomography (LACT). Standard diffusion models achieve high-quality image reconstruction but require a large number of sampling steps during inference, resulting in substantial computational overhead. Although skip-sampling strategies have been proposed to improve efficiency, they often lead to loss of fine structural details. To address this issue, we propose a prior information embedding and wavelet feature fusion fast sampling diffusion model for LACT reconstruction. The PWD enables efficient sampling while preserving reconstruction fidelity in LACT, and effectively mitigates the degradation typically introduced by skip-sampling. Specifically, during the training phase, PWD maps the distribution of LACT images to that of fully sampled target images, enabling the model to learn structural correspondences between them. During inference, the LACT image serves as an explicit prior to guide the sampling trajectory, allowing for high-quality reconstruction with significantly fewer steps. In addition, PWD performs multi-scale feature fusion in the wavelet domain, effectively enhancing the reconstruction of fine details by leveraging both low-frequency and high-frequency information. Quantitative and qualitative evaluations on clinical dental arch CBCT and periapical datasets demonstrate that PWD outperforms existing methods under the same sampling condition. Using only 50 sampling steps, PWD achieves at least 1.7 dB improvement in PSNR and 10% gain in SSIM.

eess.IV↗

arXiv2025

Dual Semantic-Aware Network for Noise Suppressed Ultrasound Video Segmentation

Ultrasound imaging is a prevalent diagnostic tool known for its simplicity and non-invasiveness. However, its inherent characteristics often introduce substantial noise, posing considerable challenges for automated lesion or organ segmentation in ultrasound video sequences. To address these limitations, we propose the Dual Semantic-Aware Network (DSANet), a novel framework designed to enhance noise robustness in ultrasound video segmentation by fostering mutual semantic awareness between local and global features. Specifically, we introduce an Adjacent-Frame Semantic-Aware (AFSA) module, which constructs a channel-wise similarity matrix to guide feature fusion across adjacent frames, effectively mitigating the impact of random noise without relying on pixel-level relationships. Additionally, we propose a Local-and-Global Semantic-Aware (LGSA) module that reorganizes and fuses temporal unconditional local features, which capture spatial details independently at each frame, with conditional global features that incorporate temporal context from adjacent frames. This integration facilitates multi-level semantic representation, significantly improving the model's resilience to noise interference. Extensive evaluations on four benchmark datasets demonstrate that DSANet substantially outperforms state-of-the-art methods in segmentation accuracy. Moreover, since our model avoids pixel-level feature dependencies, it achieves significantly higher inference FPS than video-based methods, and even surpasses some image-based models. Code can be found in \href{https://github.com/ZhouL2001/DSANet}{DSANet}

cs.CV↗

arXiv2025

PaddleOCR 3.0 Technical Report

This technical report introduces PaddleOCR 3.0, an Apache-licensed open-source toolkit for OCR and document parsing. To address the growing demand for document understanding in the era of large language models, PaddleOCR 3.0 presents three major solutions: (1) PP-OCRv5 for multilingual text recognition, (2) PP-StructureV3 for hierarchical document parsing, and (3) PP-ChatOCRv4 for key information extraction. Compared to mainstream vision-language models (VLMs), these models with fewer than 100 million parameters achieve competitive accuracy and efficiency, rivaling billion-parameter VLMs. In addition to offering a high-quality OCR model library, PaddleOCR 3.0 provides efficient tools for training, inference, and deployment, supports heterogeneous hardware acceleration, and enables developers to easily build intelligent document applications.

cs.CV↗

arXiv2025

Exploring Task Performance with Interpretable Models via Sparse Auto-Encoders

Large Language Models (LLMs) are traditionally viewed as black-box algorithms, therefore reducing trustworthiness and obscuring potential approaches to increasing performance on downstream tasks. In this work, we apply an effective LLM decomposition method using a dictionary-learning approach with sparse autoencoders. This helps extract monosemantic features from polysemantic LLM neurons. Remarkably, our work identifies model-internal misunderstanding, allowing the automatic reformulation of the prompts with additional annotations to improve the interpretation by LLMs. Moreover, this approach demonstrates a significant performance improvement in downstream tasks, such as mathematical reasoning and metaphor detection.

cs.CL↗

arXiv2025

CycleVAR: Repurposing Autoregressive Model for Unsupervised One-Step Image Translation

The current conditional autoregressive image generation methods have shown promising results, yet their potential remains largely unexplored in the practical unsupervised image translation domain, which operates without explicit cross-domain correspondences. A critical limitation stems from the discrete quantization inherent in traditional Vector Quantization-based frameworks, which disrupts gradient flow between the Variational Autoencoder decoder and causal Transformer, impeding end-to-end optimization during adversarial training in image space. To tackle this issue, we propose using Softmax Relaxed Quantization, a novel approach that reformulates codebook selection as a continuous probability mixing process via Softmax, thereby preserving gradient propagation. Building upon this differentiable foundation, we introduce CycleVAR, which reformulates image-to-image translation as image-conditional visual autoregressive generation by injecting multi-scale source image tokens as contextual prompts, analogous to prefix-based conditioning in language models. CycleVAR exploits two modes to generate the target image tokens, including (1) serial multi-step generation, enabling iterative refinement across scales, and (2) parallel one-step generation synthesizing all resolution outputs in a single forward pass. Experimental findings indicate that the parallel one-step generation mode attains superior translation quality with quicker inference speed than the serial multi-step mode in unsupervised scenarios. Furthermore, both quantitative and qualitative results indicate that CycleVAR surpasses previous state-of-the-art unsupervised image translation models, \textit{e}.\textit{g}., CycleGAN-Turbo.

cs.CV↗

arXiv2025

StereoINR: Cross-View Geometry Consistent Stereo Super Resolution with Implicit Neural Representation

Stereo image super-resolution (SSR) aims to enhance high-resolution details by leveraging information from stereo image pairs. However, existing stereo super-resolution (SSR) upsampling methods (e.g., pixel shuffle) often overlook cross-view geometric consistency and are limited to fixed-scale upsampling. The key issue is that previous upsampling methods use convolution to independently process deep features of different views, lacking cross-view and non-local information perception, making it difficult to select beneficial information from multi-view scenes adaptively. In this work, we propose Stereo Implicit Neural Representation (StereoINR), which innovatively models stereo image pairs as continuous implicit representations. This continuous representation breaks through the scale limitations, providing a unified solution for arbitrary-scale stereo super-resolution reconstruction of left-right views. Furthermore, by incorporating spatial warping and cross-attention mechanisms, StereoINR enables effective cross-view information fusion and achieves significant improvements in pixel-level geometric consistency. Extensive experiments across multiple datasets show that StereoINR outperforms out-of-training-distribution scale upsampling and matches state-of-the-art SSR methods within training-distribution scales.

eess.IV↗

arXiv2025

Benchmarking Generalizable Bimanual Manipulation: RoboTwin Dual-Arm Collaboration Challenge at CVPR 2025 MEIS Workshop

Embodied Artificial Intelligence (Embodied AI) is an emerging frontier in robotics, driven by the need for autonomous systems that can perceive, reason, and act in complex physical environments. While single-arm systems have shown strong task performance, collaborative dual-arm systems are essential for handling more intricate tasks involving rigid, deformable, and tactile-sensitive objects. To advance this goal, we launched the RoboTwin Dual-Arm Collaboration Challenge at the 2nd MEIS Workshop, CVPR 2025. Built on the RoboTwin Simulation platform (1.0 and 2.0) and the AgileX COBOT-Magic Robot platform, the competition consisted of three stages: Simulation Round 1, Simulation Round 2, and a final Real-World Round. Participants totally tackled 17 dual-arm manipulation tasks, covering rigid, deformable, and tactile-based scenarios. The challenge attracted 64 global teams and over 400 participants, producing top-performing solutions like SEM and AnchorDP3 and generating valuable insights into generalizable bimanual policy learning. This report outlines the competition setup, task design, evaluation methodology, key findings and future direction, aiming to support future research on robust and generalizable bimanual manipulation policies. The Challenge Webpage is available at https://robotwin-benchmark.github.io/cvpr-2025-challenge/.

cs.RO↗

arXiv2025

Cascaded frequency conversion of highly charged femtosecond spatiotemporal optical vortices

The degrees of freedom inherent in spatiotemporal optical vortices (STOV's) afford intriguing opportunties to manipulate complex light fields for broad applications such as optical communication, light-matter interactions, particle manipulation, quantum optics, and electron acceleration in the relativistic regime. Unlike previous studies examining the second harmonic generation (SHG) of STOV's having an input topological charge (TC) l^{\left(ω\right)}=1, here we experimentally demonstrate cascaded second and third harmonic generation of STOV's to achieve more diverse wavelength selectivity via sum frequency generation (SFG), achieving unprecedented TC values up to l^{\left(ω\right)}=40. The large TC values are attributed to second (third) harmonic generation of an incident beam satisfying l^{\left(2ω\right)}=2l^{\left(ω\right)} (l^{\left(3ω\right)}=3l^{\left(ω\right)}). What is more, the wavelength of the generated STOV field was found to be tunable by controlling the position of the phase singularity in the frequency domain, and optimizing the nonlinear phase matching condition. Our experimental measurements extend the principle of conservation of the spatiotemporal topological charge to general nonlinear optical parametric processes, suggesting a fundamental approach to the production of STOV fields of arbitrarily large TC and at arbitrary visible wavelengths, and beyond (e.g., the ultraviolet range).

physics.optics↗

arXiv2025

PP-DocBee: Improving Multimodal Document Understanding Through a Bag of Tricks

With the rapid advancement of digitalization, various document images are being applied more extensively in production and daily life, and there is an increasingly urgent need for fast and accurate parsing of the content in document images. Therefore, this report presents PP-DocBee, a novel multimodal large language model designed for end-to-end document image understanding. First, we develop a data synthesis strategy tailored to document scenarios in which we build a diverse dataset to improve the model generalization. Then, we apply a few training techniques, including dynamic proportional sampling, data preprocessing, and OCR postprocessing strategies. Extensive evaluations demonstrate the superior performance of PP-DocBee, achieving state-of-the-art results on English document understanding benchmarks and even outperforming existing open source and commercial models in Chinese document understanding. The source code and pre-trained models are publicly available at \href{https://github.com/PaddlePaddle/PaddleMIX}{https://github.com/PaddlePaddle/PaddleMIX}.

cs.CV↗

arXiv2025

PP-DocBee2: Improved Baselines with Efficient Data for Multimodal Document Understanding

This report introduces PP-DocBee2, an advanced version of the PP-DocBee, designed to enhance multimodal document understanding. Built on a large multimodal model architecture, PP-DocBee2 addresses the limitations of its predecessor through key technological improvements, including enhanced synthetic data quality, improved visual feature fusion strategy, and optimized inference methodologies. These enhancements yield an $11.4\%$ performance boost on internal benchmarks for Chinese business documents, and reduce inference latency by $73.0\%$ to the vanilla version. A key innovation of our work is a data quality optimization strategy for multimodal document tasks. By employing a large-scale multimodal pre-trained model to evaluate data, we apply a novel statistical criterion to filter outliers, ensuring high-quality training data. Inspired by insights into underutilized intermediate features in multimodal models, we enhance the ViT representational capacity by decomposing it into layers and applying a novel feature fusion strategy to improve complex reasoning. The source code and pre-trained model are available at \href{https://github.com/PaddlePaddle/PaddleMIX}{https://github.com/PaddlePaddle/PaddleMIX}.

cs.CV↗

arXiv2025

Exceedingly large in-plane critical field of finite-momentum pairing state in bulk superlattices

Magnetic flux profoundly influences the phase factor of charge particles, leading to exotic quantum phenomena. A recent example is that the orbital effect of magnetic field could induce finite-momentum pairing state in nanoflakes, which offers a new pathway to realize the spatially modulated superconductivity distinct from the Fulde-Ferrell-Larkin-Ovchinnikov (FFLO) state induced by Zeeman effect. However, whether such intriguing state can exist in the bulk materials under extremely large magnetic field remains elusive. Here we report the orbital effect induced finite-momentum pairing state with exceedingly large in-plane critical field in a bulk superconducting superlattice. Remarkably, the in-plane critical field shows a pronounced upturn behavior, exceeding eight times the Pauli limit which is comparable to monolayer Ising superconductor. Under high in-plane magnetic fields, significant anisotropic transport behavior between the interlayer and intralayer directions is detected, highlighting the critical role of suppressed interlayer coherence in the orbital effect induced finite-momentum pairing state. Crucially, this finite-momentum pairing state remains robust against moderate disorder. Our findings suggest that van der Waals superlattices, with strong Ising spin-orbit coupling and tunable interlayer coherence, offer new avenues for constructing and modulating unconventional superconducting states.

cond-mat.supr-con↗

arXiv2025

PunchBench: Benchmarking MLLMs in Multimodal Punchline Comprehension

Multimodal punchlines, which involve humor or sarcasm conveyed in image-caption pairs, are a popular way of communication on online multimedia platforms. With the rapid development of multimodal large language models (MLLMs), it is essential to assess their ability to effectively comprehend these punchlines. However, existing benchmarks on punchline comprehension suffer from three major limitations: 1) language shortcuts that allow models to solely rely on text, 2) lack of question diversity, and 3) narrow focus on a specific domain of multimodal content (e.g., cartoon). To address these limitations, we introduce a multimodal \textbf{Punch}line comprehension \textbf{Bench}mark, named \textbf{PunchBench}, which is tailored for accurate and comprehensive evaluation of punchline comprehension. To enhance the evaluation accuracy, we generate synonymous and antonymous captions by modifying original captions, which mitigates the impact of shortcuts in the captions. To provide a comprehensive evaluation, PunchBench incorporates diverse question formats and image-captions from various domains. On this basis, we conduct extensive evaluations and reveal a significant gap between state-of-the-art MLLMs and humans in punchline comprehension. To improve punchline comprehension, we propose Simple-to-Complex Chain-of-Question (SC-CoQ) strategy, enabling the models to incrementally address complicated questions by first mastering simple ones. SC-CoQ effectively enhances the performance of various MLLMs on PunchBench, surpassing in-context learning and chain-of-thought.

cs.CV↗

arXiv2025

VideoCap-R1: Enhancing MLLMs for Video Captioning via Structured Thinking

While recent advances in reinforcement learning have significantly enhanced reasoning capabilities in large language models (LLMs), these techniques remain underexplored in multi-modal LLMs for video captioning. This paper presents the first systematic investigation of GRPO-based RL post-training for video MLLMs, with the goal of enhancing video MLLMs' capability of describing actions in videos. Specifically, we develop the VideoCap-R1, which is prompted to first perform structured thinking that analyzes video subjects with their attributes and actions before generating complete captions, supported by two specialized reward mechanisms: a LLM-free think scorer evaluating the structured thinking quality and a LLM-assisted caption scorer assessing the output quality. The RL training framework effectively establishes the connection between structured reasoning and comprehensive description generation, enabling the model to produce captions with more accurate actions. Our experiments demonstrate that VideoCap-R1 achieves substantial improvements over the Qwen2VL-7B baseline using limited samples (1.5k) across multiple video caption benchmarks (DREAM1K: +4.4 event F1, VDC: +4.2 Acc, CAREBENCH: +3.1 action F1, +6.9 object F1) while consistently outperforming the SFT-trained counterparts, confirming GRPO's superiority in enhancing MLLMs' captioning capabilities.

cs.CV↗

arXiv2025

Optical Photometric Monitoring of the Blazar OT 355 and Local Standard Stars' Calibration

OT 355 (4FGL J1734.3 + 3858) is a relatively rarely studied but highly variable, moderate-redshift (z = 0.975) flat-spectrum radio quasar (blazar). With this work, we aim to study its optical variability on different timescales, which can help us to better understand the physical processes in relativistic jets operating in blazar-type active galactic nuclei. OT 355 was observed in four colors (BVRI) during 41 nights between 2017 and 2023 using three 1 and 2 m class telescopes. The object was also monitored on intra-night timescales, for about 100 h in total. In addition, secondary standard stars in the field of OT 355 were calibrated in order to facilitate future photometric studies. We detected significant intra-night and night-to-night variations of up to 0.5 mag. Variability characteristics, color changes, and a possible ``rms-flux'' relation were studied and discussed. Using simple arguments, we show that a negative ``rms-flux'' relation should be expected if many independent processes/regions drive the short-term variability via Doppler factor changes, which is not observed in this and other cases. This finding raises arguments for the idea that more complex multiplicative processes are responsible for blazar variability. Studying blazar variability, especially on the shortest possible timescales, can help to estimate the strength and geometry of their magnetic fields, the linear sizes of the emitting regions, and other aspects, which may be of importance for constraining and modeling blazars' emitting mechanisms.

astro-ph.HE↗

arXiv2025

DeepRTL2: A Versatile Model for RTL-Related Tasks

The integration of large language models (LLMs) into electronic design automation (EDA) has significantly advanced the field, offering transformative benefits, particularly in register transfer level (RTL) code generation and understanding. While previous studies have demonstrated the efficacy of fine-tuning LLMs for these generation-based tasks, embedding-based tasks, which are equally critical to EDA workflows, have been largely overlooked. These tasks, including natural language code search, RTL code functionality equivalence checking, and performance prediction, are essential for accelerating and optimizing the hardware design process. To address this gap, we present DeepRTL2, a family of versatile LLMs that unifies both generation- and embedding-based tasks related to RTL. By simultaneously tackling a broad range of tasks, DeepRTL2 represents the first model to provide a comprehensive solution to the diverse challenges in EDA. Through extensive experiments, we show that DeepRTL2 achieves state-of-the-art performance across all evaluated tasks.

cs.AR↗

arXiv2025

A statistically consistent measure of semantic uncertainty using Language Models

To address the challenge of quantifying uncertainty in the outputs generated by language models, we propose a novel measure of semantic uncertainty, semantic spectral entropy, that is statistically consistent under mild assumptions. This measure is implemented through a straightforward algorithm that relies solely on standard, pretrained language models, without requiring access to the internal generation process. Our approach imposes minimal constraints on the choice of language models, making it broadly applicable across different architectures and settings. Through comprehensive simulation studies, we demonstrate that the proposed method yields an accurate and robust estimate of semantic uncertainty, even in the presence of the inherent randomness characteristic of generative language model outputs.

cs.CL↗

arXiv2025

SpaceR: Reinforcing MLLMs in Video Spatial Reasoning

Video spatial reasoning, which involves inferring the underlying spatial structure from observed video frames, poses a significant challenge for existing Multimodal Large Language Models (MLLMs). This limitation stems primarily from 1) the absence of high-quality datasets for this task, and 2) the lack of effective training strategies to develop spatial reasoning capabilities. Motivated by the success of Reinforcement Learning with Verifiable Reward (RLVR) in unlocking LLM reasoning abilities, this work aims to improve MLLMs in video spatial reasoning through the RLVR paradigm. To this end, we introduce the $\textbf{SpaceR}$ framework. First, we present $\textbf{SpaceR-151k}$, a dataset with 91k questions spanning diverse spatial reasoning scenarios with verifiable answers, and 60k samples for maintaining general multimodal understanding. Second, we propose $\textbf{Spatially-Guided RLVR (SG-RLVR)}$, a novel reinforcement learning approach that extends Group Relative Policy Optimization (GRPO) with a novel map imagination mechanism, which encourages the model to infer spatial layouts in the thinking process, thereby facilitating more effective spatial reasoning. Extensive experiments demonstrate that SpaceR achieves state-of-the-art performance on spatial reasoning benchmarks (e.g., VSI-Bench, STI-Bench, and SPAR-Bench), while maintaining competitive results on video understanding benchmarks (e.g., Video-MME, TempCompass, and LongVideoBench). Remarkably, SpaceR surpasses the advanced GPT-4o by 11.6\% accuracy on VSI-Bench and is on par with the leading proprietary model Gemini-2.0-Flash, highlighting the effectiveness of our SpaceR-151k dataset and SG-RLVR in reinforcing spatial reasoning ability of MLLMs. Code, model, and dataset are available at https://github.com/OuyangKun10/SpaceR.

cs.CV↗

arXiv2025

D3: Diversity, Difficulty, and Dependability-Aware Data Selection for Sample-Efficient LLM Instruction Tuning

Recent advancements in instruction tuning for large language models (LLMs) suggest that a small, high-quality dataset can significantly equip LLMs with instruction-following capabilities, outperforming large datasets often burdened by quality and redundancy issues. However, the challenge lies in automatically identifying valuable subsets from large datasets to boost both the effectiveness and efficiency of instruction tuning. In this paper, we first establish data selection criteria based on three distinct aspects of data value: diversity, difficulty, and dependability, and then propose the D3 method comprising two key steps of scoring and selection. Specifically, in the scoring step, we define the diversity function to measure sample distinctiveness and introduce the uncertainty-based prediction difficulty to evaluate sample difficulty by mitigating the interference of context-oriented generation diversity. Additionally, we integrate an external LLM for dependability assessment. In the selection step, we formulate the D3 weighted coreset objective, which jointly optimizes three aspects of data value to solve for the most valuable subset. The two steps of D3 can iterate multiple rounds, incorporating feedback to refine the selection focus adaptively. Experiments on both public datasets and the real-world Taobao Live application demonstrate the effectiveness of D3 in endowing LLMs with competitive or even superior instruction-following capabilities using less than 10\% of the entire dataset.

cs.LG↗