arXiv subjects

Liuyue Xie

Publications and source records attributed to Liuyue Xie.

8 recordsLinked to original sources

arXiv2026

Escaping Confidence Trap: Evolutionary Decoding for Mathematical Reasoning in Diffusion LLMs

Diffusion large language models (dLLMs) have emerged as a promising alternative to autoregressive LLMs, offering efficient generation through block-wise progressive unmasking. However, their strong general-purpose performance does not necessarily translate into reliable mathematical reasoning, where correctness depends on preserving coherent numerical-symbolic reasoning trajectories. In this work, we analyze the decoding trajectories of LLaDA 2.0 and identify a recurring diffusion confidence trap: local token confidence can become misaligned with global reasoning correctness during progressive block decoding. Our analysis reveals two representative failure regimes: sampling-sensitive failures, where correct paths exist but are unstable, and sampling-consistent failures, where repeated sampling converges to repetitive high-confidence but incorrect continuations. Motivated by this observation, we propose Evolutionary Decoding, a training-free test-time scaling framework that views diffusion decoding as an evolutionary process over candidate reasoning states. The framework combines step-wise selection, which preserves useful numerical-symbolic signals and suppresses repetitive patterns, with block-wise mutation, which introduces structured alternatives to escape incorrect high-confidence basins. Experiments on multiple benchmarks show that Evolutionary Decoding improves LLaDA 2.0 over confidence-based decoding, leading to more reliable mathematical reasoning.

cs.AI↗

arXiv2026

Unified Spherical Frontend: Learning Rotation-Equivariant Representations of Spherical Images from Any Camera

Modern perception increasingly relies on fisheye, panoramic, and other wide field-of-view (FoV) cameras, yet most pipelines still apply planar CNNs designed for pinhole imagery on 2D grids, where pixel-space neighborhoods misrepresent physical adjacency and models are sensitive to global rotations. Traditional spherical CNNs partially address this mismatch but require costly spherical harmonic transform that constrains resolution and efficiency. We present Unified Spherical Frontend (USF), a distortion-free lens-agnostic framework that transforms images from any calibrated camera onto the unit sphere via ray-direction correspondences, and performs spherical resampling, convolution, and pooling canonically in the spatial domain. USF is modular: projection, location sampling, value interpolation, and resolution control are fully decoupled. Its configurable distance-only convolution kernels offer rotation-equivariance, mirroring translation-equivariance in planar CNNs while avoiding harmonic transforms entirely. We compare multiple standard planar backbones with their spherical counterparts across classification, detection, and segmentation tasks on synthetic (Spherical MNIST) and real-world (PANDORA, Stanford 2D-3D-S) datasets, and stress-test robustness to extreme lens distortions, varying FoV, and arbitrary rotations. USF scales efficiently to high-resolution spherical imagery and maintains less than 1% performance drop under random test-time rotations without training-time rotational augmentation, and enables zero-shot generalization to any unseen (wide-FoV) lenses with minimal performance degradation.

cs.CV↗

arXiv2025

MAVERIX: Multimodal Audio-Visual Evaluation and Recognition IndeX

We introduce MAVERIX (Multimodal audiovisual Evaluation and Recognition IndeX), a unified benchmark to probe the video understanding in multimodal LLMs, encompassing video, audio, text inputs with human performance baselines. Although recent advancements in models with vision and audio understanding capabilities have shown substantial progress, the field lacks a standardized evaluation framework to thoroughly assess their cross-modality comprehension performance. MAVERIX curates 2,556 questions from 700 videos, in the form of both multiple-choice and open-ended formats, explicitly designed to evaluate multimodal models through questions that necessitate tight integration of video and audio information, spanning a broad spectrum of agentic scenarios. MAVERIX uniquely provides models with audiovisual questions, closely mimicking the multimodal perceptual experiences available to humans during inference and decision-making processes. To our knowledge, MAVERIX is the first benchmark aimed explicitly at assessing comprehensive audiovisual integration in such granularity. Experiments with state-of-the-art models, including Qwen 2.5 Omni and Gemini 2.5 Flash-Lite, show performance around 64% accuracy, while human experts reach near-ceiling performance of 92.8%, exposing a substantial gap to human-level comprehension. With standardized evaluation protocols, a rigorously annotated pipeline, and a public toolkit, MAVERIX establishes a challenging testbed for advancing audiovisual multimodal intelligence.

cs.MM↗

arXiv2025

Gaussian Splatting Lucas-Kanade

Gaussian Splatting and its dynamic extensions are effective for reconstructing 3D scenes from 2D images when there is significant camera movement to facilitate motion parallax and when scene objects remain relatively static. However, in many real-world scenarios, these conditions are not met. As a consequence, data-driven semantic and geometric priors have been favored as regularizers, despite their bias toward training data and their neglect of broader movement dynamics. Departing from this practice, we propose a novel analytical approach that adapts the classical Lucas-Kanade method to dynamic Gaussian splatting. By leveraging the intrinsic properties of the forward warp field network, we derive an analytical velocity field that, through time integration, facilitates accurate scene flow computation. This enables the precise enforcement of motion constraints on warp fields, thus constraining both 2D motion and 3D positions of the Gaussians. Our method excels in reconstructing highly dynamic scenes with minimal camera movement, as demonstrated through experiments on both synthetic and real-world scenes.

cs.CV↗

arXiv2025

AlignDiff: Learning Physically-Grounded Camera Alignment via Diffusion

Accurate camera calibration is a fundamental task for 3D perception, especially when dealing with real-world, in-the-wild environments where complex optical distortions are common. Existing methods often rely on pre-rectified images or calibration patterns, which limits their applicability and flexibility. In this work, we introduce a novel framework that addresses these challenges by jointly modeling camera intrinsic and extrinsic parameters using a generic ray camera model. Unlike previous approaches, AlignDiff shifts focus from semantic to geometric features, enabling more accurate modeling of local distortions. We propose AlignDiff, a diffusion model conditioned on geometric priors, enabling the simultaneous estimation of camera distortions and scene geometry. To enhance distortion prediction, we incorporate edge-aware attention, focusing the model on geometric features around image edges, rather than semantic content. Furthermore, to enhance generalizability to real-world captures, we incorporate a large database of ray-traced lenses containing over three thousand samples. This database characterizes the distortion inherent in a diverse variety of lens forms. Our experiments demonstrate that the proposed method significantly reduces the angular error of estimated ray bundles by ~8.2 degrees and overall calibration accuracy, outperforming existing approaches on challenging, real-world datasets.

cs.CV↗

arXiv2024

Through the Curved Cover: Synthesizing Cover Aberrated Scenes with Refractive Field

Recent extended reality headsets and field robots have adopted covers to protect the front-facing cameras from environmental hazards and falls. The surface irregularities on the cover can lead to optical aberrations like blurring and non-parametric distortions. Novel view synthesis methods like NeRF and 3D Gaussian Splatting are ill-equipped to synthesize from sequences with optical aberrations. To address this challenge, we introduce SynthCover to enable novel view synthesis through protective covers for downstream extended reality applications. SynthCover employs a Refractive Field that estimates the cover's geometry, enabling precise analytical calculation of refracted rays. Experiments on synthetic and real-world scenes demonstrate our method's ability to accurately model scenes viewed through protective covers, achieving a significant improvement in rendering quality compared to prior methods. We also show that the model can adjust well to various cover geometries with synthetic sequences captured with covers of different surface curvatures. To motivate further studies on this problem, we provide the benchmarked dataset containing real and synthetic walkable scenes captured with protective cover optical aberrations.

cs.CV↗

arXiv2023

Component Segmentation of Engineering Drawings Using Graph Convolutional Networks

We present a data-driven framework to automate the vectorization and machine interpretation of 2D engineering part drawings. In industrial settings, most manufacturing engineers still rely on manual reads to identify the topological and manufacturing requirements from drawings submitted by designers. The interpretation process is laborious and time-consuming, which severely inhibits the efficiency of part quotation and manufacturing tasks. While recent advances in image-based computer vision methods have demonstrated great potential in interpreting natural images through semantic segmentation approaches, the application of such methods in parsing engineering technical drawings into semantically accurate components remains a significant challenge. The severe pixel sparsity in engineering drawings also restricts the effective featurization of image-based data-driven methods. To overcome these challenges, we propose a deep learning based framework that predicts the semantic type of each vectorized component. Taking a raster image as input, we vectorize all components through thinning, stroke tracing, and cubic bezier fitting. Then a graph of such components is generated based on the connectivity between the components. Finally, a graph convolutional neural network is trained on this graph data to identify the semantic type of each component. We test our framework in the context of semantic segmentation of text, dimension and, contour components in engineering drawings. Results show that our method yields the best performance compared to recent image, and graph-based segmentation methods.

cs.CV↗

arXiv2020

Multi-Resolution Graph Neural Network for Large-Scale Pointcloud Segmentation

In this paper, we propose a multi-resolution deep-learning architecture to semantically segment dense large-scale pointclouds. Dense pointcloud data require a computationally expensive feature encoding process before semantic segmentation. Previous work has used different approaches to drastically downsample from the original pointcloud so common computing hardware can be utilized. While these approaches can relieve the computation burden to some extent, they are still limited in their processing capability for multiple scans. We present MuGNet, a memory-efficient, end-to-end graph neural network framework to perform semantic segmentation on large-scale pointclouds. We reduce the computation demand by utilizing a graph neural network on the preformed pointcloud graphs and retain the precision of the segmentation with a bidirectional network that fuses feature embedding at different resolutions. Our framework has been validated on benchmark datasets including Stanford Large-Scale 3D Indoor Spaces Dataset(S3DIS) and Virtual KITTI Dataset. We demonstrate that our framework can process up to 45 room scans at once on a single 11 GB GPU while still surpassing other graph-based solutions for segmentation on S3DIS with an 88.5\% (+3\%) overall accuracy and 69.8\% (+7.7\%) mIOU accuracy.

cs.CV↗