arXiv subjects

Yan Lu

Publications and source records attributed to Yan Lu.

At least 127 records · Page 7Linked to original sources

arXiv2025

DLF: Extreme Image Compression with Dual-generative Latent Fusion

Recent studies in extreme image compression have achieved remarkable performance by compressing the tokens from generative tokenizers. However, these methods often prioritize clustering common semantics within the dataset, while overlooking the diverse details of individual objects. Consequently, this results in suboptimal reconstruction fidelity, especially at low bitrates. To address this issue, we introduce a Dual-generative Latent Fusion (DLF) paradigm. DLF decomposes the latent into semantic and detail elements, compressing them through two distinct branches. The semantic branch clusters high-level information into compact tokens, while the detail branch encodes perceptually critical details to enhance the overall fidelity. Additionally, we propose a cross-branch interactive design to reduce redundancy between the two branches, thereby minimizing the overall bit cost. Experimental results demonstrate the impressive reconstruction quality of DLF even below 0.01 bits per pixel (bpp). On the CLIC2020 test set, our method achieves bitrate savings of up to 27.93% on LPIPS and 53.55% on DISTS compared to MS-ILLM. Furthermore, DLF surpasses recent diffusion-based codecs in visual fidelity while maintaining a comparable level of generative realism. Project: https://dlfcodec.github.io/

cs.CV↗

arXiv2025

Bitrate-Controlled Diffusion for Disentangling Motion and Content in Video

We propose a novel and general framework to disentangle video data into its dynamic motion and static content components. Our proposed method is a self-supervised pipeline with less assumptions and inductive biases than previous works: it utilizes a transformer-based architecture to jointly generate flexible implicit features for frame-wise motion and clip-wise content, and incorporates a low-bitrate vector quantization as an information bottleneck to promote disentanglement and form a meaningful discrete motion space. The bitrate-controlled latent motion and content are used as conditional inputs to a denoising diffusion model to facilitate self-supervised representation learning. We validate our disentangled representation learning framework on real-world talking head videos with motion transfer and auto-regressive motion generation tasks. Furthermore, we also show that our method can generalize to other types of video data, such as pixel sprites of 2D cartoon characters. Our work presents a new perspective on self-supervised learning of disentangled video representations, contributing to the broader field of video analysis and generation.

cs.CV↗

arXiv2025

FELLE: Autoregressive Speech Synthesis with Token-Wise Coarse-to-Fine Flow Matching

To advance continuous-valued token modeling and temporal-coherence enforcement, we propose FELLE, an autoregressive model that integrates language modeling with token-wise flow matching. By leveraging the autoregressive nature of language models and the generative efficacy of flow matching, FELLE effectively predicts continuous-valued tokens (mel-spectrograms). For each continuous-valued token, FELLE modifies the general prior distribution in flow matching by incorporating information from the previous step, improving coherence and stability. Furthermore, to enhance synthesis quality, FELLE introduces a coarse-to-fine flow-matching mechanism, generating continuous-valued tokens hierarchically, conditioned on the language model's output. Experimental results demonstrate the potential of incorporating flow-matching techniques in autoregressive mel-spectrogram modeling, leading to significant improvements in TTS generation quality, as shown in https://aka.ms/felle.

cs.CL↗

arXiv2025

Refining fundamental constants with white dwarfs: machine learning informed constraints on fine-structure constant and proton-to-electron mass ratio

We explore the potential variation of two fundamental constants, the fine-structure constant $α$ and the proton-to-electron mass ratio $μ$, within the framework of modified gravity theories and finite-temperature effects. Utilising high-precision white dwarf observations from the Gaia-DR3 survey, we construct a robust mass--radius relation using a Bayesian-inspired machine learning framework. This empirical relation is rigorously compared with theoretical predictions derived from scalar-tensor gravity models and temperature-dependent equations of state. Our results demonstrate that both underlying gravitational theory and temperature substantially influence the inferred constraints on $α$ and $μ$. We obtain the strongest constraints as $|Δα/α|=2.10^{+32.56}_{-39.26}\times10^{-7}$ and $|Δμ/μ|=1.61^{+37.16}_{-34.67}\times10^{-7}$ for modified gravity parameter $γ\simeq -3.69\times10^{13}\,\mathrm{cm}^2$, while for the finite temperature case, these are $|Δα/α|=1.60^{+37.31}_{-35.42}\times10^{-7}$ and $|Δμ/μ|=1.23^{+37.02}_{-35.71}\times10^{-7}$ for $T \simeq 1.1 \times 10^7\rm\, K$. These findings yield tighter constraints than those reported in earlier studies and underscore the critical roles of gravitational and thermal physics in testing the constancy of fundamental parameters.

astro-ph.SR↗

arXiv2025

TrInk: Ink Generation with Transformer Network

In this paper, we propose TrInk, a Transformer-based model for ink generation, which effectively captures global dependencies. To better facilitate the alignment between the input text and generated stroke points, we introduce scaled positional embeddings and a Gaussian memory mask in the cross-attention module. Additionally, we design both subjective and objective evaluation pipelines to comprehensively assess the legibility and style consistency of the generated handwriting. Experiments demonstrate that our Transformer-based model achieves a 35.56\% reduction in character error rate (CER) and an 29.66% reduction in word error rate (WER) on the IAM-OnDB dataset compared to previous methods. We provide an demo page with handwriting samples from TrInk and baseline models at: https://akahello-a11y.github.io/trink-demo/

cs.CL↗

arXiv2025

Interleaved Speech-Text Language Models for Simple Streaming Text-to-Speech Synthesis

This paper introduces Interleaved Speech-Text Language Model (IST-LM) for zero-shot streaming Text-to-Speech (TTS). Unlike many previous approaches, IST-LM is directly trained on interleaved sequences of text and speech tokens with a fixed ratio, eliminating the need for additional efforts like forced alignment or complex designs. The ratio of text chunk size to speech chunk size is crucial for the performance of IST-LM. To explore this, we conducted a comprehensive series of statistical analyses on the training data and performed correlation analysis with the final performance, uncovering several key factors: 1) the distance between speech tokens and their corresponding text tokens, 2) the number of future text tokens accessible to each speech token, and 3) the frequency of speech tokens precedes their corresponding text tokens. Experimental results demonstrate how to achieve an optimal streaming TTS system with a limited performance gap compared to its non-streaming counterpart. IST-LM is conceptually simple and empirically powerful, enabling streaming TTS with minimal overhead while largely preserving performance, and offering broad potential for integration with real-time text streams from large language models.

eess.AS↗

arXiv2025

Pseudo-Autoregressive Neural Codec Language Models for Efficient Zero-Shot Text-to-Speech Synthesis

Recent zero-shot text-to-speech (TTS) systems face a common dilemma: autoregressive (AR) models suffer from slow generation and lack duration controllability, while non-autoregressive (NAR) models lack temporal modeling and typically require complex designs. In this paper, we introduce a novel pseudo-autoregressive (PAR) codec language modeling approach that unifies AR and NAR modeling. Combining explicit temporal modeling from AR with parallel generation from NAR, PAR generates dynamic-length spans at fixed time steps. Building on PAR, we propose PALLE, a two-stage TTS system that leverages PAR for initial generation followed by NAR refinement. In the first stage, PAR progressively generates speech tokens along the time dimension, with each step predicting all positions in parallel but only retaining the left-most span. In the second stage, low-confidence tokens are iteratively refined in parallel, leveraging the global contextual information. Experiments demonstrate that PALLE, trained on LibriTTS, outperforms state-of-the-art systems trained on large-scale data, including F5-TTS, E2-TTS, and MaskGCT, on the LibriSpeech test-clean set in terms of speech quality, speaker similarity, and intelligibility, while achieving up to ten times faster inference speed. Audio samples are available at https://microsoft.com/research/project/vall-e-x/palle.

eess.AS↗

arXiv2025

MLLM-based Discovery of Intrinsic Coordinates and Governing Equations from High-Dimensional Data

Discovering governing equations from scientific data is crucial for understanding the evolution of systems, and is typically framed as a search problem within a candidate equation space. However, the high-dimensional nature of dynamical systems leads to an exponentially expanding equation space, making the search process extremely challenging. The visual perception and pre-trained scientific knowledge of multimodal large language models (MLLM) hold promise for providing effective navigation in high-dimensional equation spaces. In this paper, we propose a zero-shot method based on MLLM for automatically discovering physical coordinates and governing equations from high-dimensional data. Specifically, we design a series of enhanced visual prompts for MLLM to enhance its spatial perception. In addition, MLLM's domain knowledge is employed to navigate the search process within the equation space. Quantitative and qualitative evaluations on two representative types of systems demonstrate that the proposed method effectively discovers the physical coordinates and equations from both simulated and real experimental data, with long-term extrapolation accuracy improved by approximately 26.96% compared to the baseline.

cs.CE↗

arXiv2025

ForenX: Towards Explainable AI-Generated Image Detection with Multimodal Large Language Models

Advances in generative models have led to AI-generated images visually indistinguishable from authentic ones. Despite numerous studies on detecting AI-generated images with classifiers, a gap persists between such methods and human cognitive forensic analysis. We present ForenX, a novel method that not only identifies the authenticity of images but also provides explanations that resonate with human thoughts. ForenX employs the powerful multimodal large language models (MLLMs) to analyze and interpret forensic cues. Furthermore, we overcome the limitations of standard MLLMs in detecting forgeries by incorporating a specialized forensic prompt that directs the MLLMs attention to forgery-indicative attributes. This approach not only enhance the generalization of forgery detection but also empowers the MLLMs to provide explanations that are accurate, relevant, and comprehensive. Additionally, we introduce ForgReason, a dataset dedicated to descriptions of forgery evidences in AI-generated images. Curated through collaboration between an LLM-based agent and a team of human annotators, this process provides refined data that further enhances our model's performance. We demonstrate that even limited manual annotations significantly improve explanation quality. We evaluate the effectiveness of ForenX on two major benchmarks. The model's explainability is verified by comprehensive subjective evaluations.

cs.CV↗

arXiv2025

UI-E2I-Synth: Advancing GUI Grounding with Large-Scale Instruction Synthesis

Recent advancements in Large Vision-Language Models are accelerating the development of Graphical User Interface (GUI) agents that utilize human-like vision perception capabilities to enhance productivity on digital devices. Compared to approaches predicated on GUI metadata, which are platform-dependent and vulnerable to implementation variations, vision-based approaches offer broader applicability. In this vision-based paradigm, the GUI instruction grounding, which maps user instruction to the location of corresponding element on the given screenshot, remains a critical challenge, particularly due to limited public training dataset and resource-intensive manual instruction data annotation. In this paper, we delve into unexplored challenges in this task including element-to-screen ratio, unbalanced element type, and implicit instruction. To address these challenges, we introduce a large-scale data synthesis pipeline UI-E2I-Synth for generating varying complex instruction datasets using GPT-4o instead of human annotators. Furthermore, we propose a new GUI instruction grounding benchmark UI-I2E-Bench, which is designed to address the limitations of existing benchmarks by incorporating diverse annotation aspects. Our model, trained on the synthesized data, achieves superior performance in GUI instruction grounding, demonstrating the advancements of proposed data synthesis pipeline. The proposed benchmark, accompanied by extensive analyses, provides practical insights for future research in GUI grounding. We will release corresponding artifacts at https://microsoft.github.io/FIVE-UI-Evol/ .

cs.HC↗

arXiv2025

High Performance Space Debris Tracking in Complex Skylight Backgrounds with a Large-Scale Dataset

With the rapid development of space exploration, space debris has attracted more attention due to its potential extreme threat, leading to the need for real-time and accurate debris tracking. However, existing methods are mainly based on traditional signal processing, which cannot effectively process the complex background and dense space debris. In this paper, we propose a deep learning-based Space Debris Tracking Network~(SDT-Net) to achieve highly accurate debris tracking. SDT-Net effectively represents the feature of debris, enhancing the efficiency and stability of end-to-end model learning. To train and evaluate this model effectively, we also produce a large-scale dataset Space Debris Tracking Dataset (SDTD) by a novel observation-based data simulation scheme. SDTD contains 18,040 video sequences with a total of 62,562 frames and covers 250,000 synthetic space debris. Extensive experiments validate the effectiveness of our model and the challenging of our dataset. Furthermore, we test our model on real data from the Antarctic Station, achieving a MOTA score of 73.2%, which demonstrates its strong transferability to real-world scenarios. Our dataset and code will be released soon.

cs.CV↗

arXiv2025

From Virtual Games to Real-World Play

We introduce RealPlay, a neural network-based real-world game engine that enables interactive video generation from user control signals. Unlike prior works focused on game-style visuals, RealPlay aims to produce photorealistic, temporally consistent video sequences that resemble real-world footage. It operates in an interactive loop: users observe a generated scene, issue a control command, and receive a short video chunk in response. To enable such realistic and responsive generation, we address key challenges including iterative chunk-wise prediction for low-latency feedback, temporal consistency across iterations, and accurate control response. RealPlay is trained on a combination of labeled game data and unlabeled real-world videos, without requiring real-world action annotations. Notably, we observe two forms of generalization: (1) control transfer-RealPlay effectively maps control signals from virtual to real-world scenarios; and (2) entity transfer-although training labels originate solely from a car racing game, RealPlay generalizes to control diverse real-world entities, including bicycles and pedestrians, beyond vehicles. Project page can be found: https://wenqsun.github.io/RealPlay/

cs.CV↗

arXiv2025

StreamMel: Real-Time Zero-shot Text-to-Speech via Interleaved Continuous Autoregressive Modeling

Recent advances in zero-shot text-to-speech (TTS) synthesis have achieved high-quality speech generation for unseen speakers, but most systems remain unsuitable for real-time applications because of their offline design. Current streaming TTS paradigms often rely on multi-stage pipelines and discrete representations, leading to increased computational cost and suboptimal system performance. In this work, we propose StreamMel, a pioneering single-stage streaming TTS framework that models continuous mel-spectrograms. By interleaving text tokens with acoustic frames, StreamMel enables low-latency, autoregressive synthesis while preserving high speaker similarity and naturalness. Experiments on LibriSpeech demonstrate that StreamMel outperforms existing streaming TTS baselines in both quality and latency. It even achieves performance comparable to offline systems while supporting efficient real-time generation, showcasing broad prospects for integration with real-time speech large language models. Audio samples are available at: https://aka.ms/StreamMel.

cs.SD↗

arXiv2025

Perfecting Depth: Uncertainty-Aware Enhancement of Metric Depth

We propose a novel two-stage framework for sensor depth enhancement, called Perfecting Depth. This framework leverages the stochastic nature of diffusion models to automatically detect unreliable depth regions while preserving geometric cues. In the first stage (stochastic estimation), the method identifies unreliable measurements and infers geometric structure by leveraging a training-inference domain gap. In the second stage (deterministic refinement), it enforces structural consistency and pixel-level accuracy using the uncertainty map derived from the first stage. By combining stochastic uncertainty modeling with deterministic refinement, our method yields dense, artifact-free depth maps with improved reliability. Experimental results demonstrate its effectiveness across diverse real-world scenarios. Furthermore, theoretical analysis, various experiments, and qualitative visualizations validate its robustness and scalability. Our framework sets a new baseline for sensor depth enhancement, with potential applications in autonomous driving, robotics, and immersive technologies.

cs.CV↗

arXiv2025

Zero-Shot Streaming Text to Speech Synthesis with Transducer and Auto-Regressive Modeling

Zero-shot streaming text-to-speech is an important research topic in human-computer interaction. Existing methods primarily use a lookahead mechanism, relying on future text to achieve natural streaming speech synthesis, which introduces high processing latency. To address this issue, we propose SMLLE, a streaming framework for generating high-quality speech frame-by-frame. SMLLE employs a Transducer to convert text into semantic tokens in real time while simultaneously obtaining duration alignment information. The combined outputs are then fed into a fully autoregressive (AR) streaming model to reconstruct mel-spectrograms. To further stabilize the generation process, we design a Delete < Bos > Mechanism that allows the AR model to access future text introducing as minimal delay as possible. Experimental results suggest that the SMLLE outperforms current streaming TTS methods and achieves comparable performance over sentence-level TTS systems. Samples are available on shy-98.github.io/SMLLE_demo_page/.

cs.LG↗

arXiv2025

Real-time Light Curve Classification Framework for the Wide Field Survey Telescope Using Modified Semi-supervised Variational Auto-Encoder

Modern time-domain astronomy will benefit from the vast data collected by survey telescopes. The 2.5 m Wide Field Survey Telescope (WFST), with its powerful capabilities, is promising to make significant contributions in the era of large sky surveys. To harness the full potential of the enormous amount of unlabeled light curve data that the WFST will collect, we have developed a semisupervised light curve classification framework. This framework showcases several unique features. First, it is optimized for classifying events based on the early phase of the light curve (three days after trigger), which can help identify interesting events early and enable efficient follow-up observations. Second, the semisupervised nature of our framework allows it to leverage valuable information from large volumes of unlabeled data, potentially bridging the gap between simulations and real observations and achieving better generalization in practical scenarios. Compared to the commonly used Recurrent Neural Network models, our framework has shown a 5.59% improvement in accuracy for early classification tasks, as well as improvements in precision and recall in almost all subclasses. Moreover, our approach provides a reconstructed light curve, along with a compact latent representation, offering a different perspective that can be used for further downstream tasks beyond classification. The code and model weights used in this work are maintained and publicly available on our GitHub repository.

astro-ph.IM↗

arXiv2025

LTM3D: Bridging Token Spaces for Conditional 3D Generation with Auto-Regressive Diffusion Framework

We present LTM3D, a Latent Token space Modeling framework for conditional 3D shape generation that integrates the strengths of diffusion and auto-regressive (AR) models. While diffusion-based methods effectively model continuous latent spaces and AR models excel at capturing inter-token dependencies, combining these paradigms for 3D shape generation remains a challenge. To address this, LTM3D features a Conditional Distribution Modeling backbone, leveraging a masked autoencoder and a diffusion model to enhance token dependency learning. Additionally, we introduce Prefix Learning, which aligns condition tokens with shape latent tokens during generation, improving flexibility across modalities. We further propose a Latent Token Reconstruction module with Reconstruction-Guided Sampling to reduce uncertainty and enhance structural fidelity in generated shapes. Our approach operates in token space, enabling support for multiple 3D representations, including signed distance fields, point clouds, meshes, and 3D Gaussian Splatting. Extensive experiments on image- and text-conditioned shape generation tasks demonstrate that LTM3D outperforms existing methods in prompt fidelity and structural accuracy while offering a generalizable framework for multi-modal, multi-representation 3D generation.

cs.CV↗

arXiv2025

Generative Latent Coding for Ultra-Low Bitrate Image and Video Compression

Most existing approaches for image and video compression perform transform coding in the pixel space to reduce redundancy. However, due to the misalignment between the pixel-space distortion and human perception, such schemes often face the difficulties in achieving both high-realism and high-fidelity at ultra-low bitrate. To solve this problem, we propose \textbf{G}enerative \textbf{L}atent \textbf{C}oding (\textbf{GLC}) models for image and video compression, termed GLC-image and GLC-Video. The transform coding of GLC is conducted in the latent space of a generative vector-quantized variational auto-encoder (VQ-VAE). Compared to the pixel-space, such a latent space offers greater sparsity, richer semantics and better alignment with human perception, and show its advantages in achieving high-realism and high-fidelity compression. To further enhance performance, we improve the hyper prior by introducing a spatial categorical hyper module in GLC-image and a spatio-temporal categorical hyper module in GLC-video. Additionally, the code-prediction-based loss function is proposed to enhance the semantic consistency. Experiments demonstrate that our scheme shows high visual quality at ultra-low bitrate for both image and video compression. For image compression, GLC-image achieves an impressive bitrate of less than $0.04$ bpp, achieving the same FID as previous SOTA model MS-ILLM while using $45\%$ fewer bitrate on the CLIC 2020 test set. For video compression, GLC-video achieves 65.3\% bitrate saving over PLVC in terms of DISTS.

eess.IV↗