arXiv subjects

Daniel Xu

Publications and source records attributed to Daniel Xu.

6 recordsLinked to original sources

arXiv2026

Meshtryoshka: Differentiable Rendering of Real-World Scenes via Mesh Rasterization

Differentiable rendering has emerged as a powerful approach for 3D reconstruction and novel view synthesis. State-of-the-art differentiable rendering methods combine a variety of custom representations of 3D geometry and appearance with specialized renderers. However, most downstream tasks in computer graphics rely on 3D meshes. While prior work has attempted differentiable rendering with mesh representations, these approaches are limited to object-centric scenes and fail to reconstruct large-scale, unbounded scenes. In this work, we introduce Meshtryoshka, a novel mesh differentiable rendering framework that combines an off-the-shelf triangle rasterizer with a 3D representation that consists of nested mesh shells which resemble a matryoshka doll. In every forward pass, the mesh shells are extracted anew from a 3D signed distance function via iso-surface extraction, and the opacities for each vertex are computed as a function of signed distance. Each mesh shell is then rasterized independently, and the final image is created via alpha compositing. Crucially, mesh vertex positions are updated only indirectly via gradients that flow through the opacity values into the signed distance function, and hence, our method is compatible with off-the-shelf mesh renderers that need not be differentiable with respect to vertex positions. On object-centric scenes, our method performs competitively with surface-based differentiable rendering techniques. Our differentiable mesh rendering method scales to unbounded, real-world 3D scenes, where it yields high-quality novel view synthesis results approaching those of state-of-the-art, non-mesh methods. Our method suggests that it may be possible to solve the differentiable rendering problem without relying on specialized renderers, only using conventional tools from the computer graphics toolbox.

cs.CV↗

arXiv2026

Neural Statistical Functions

Classical deep learning typically operates on individual cases. Despite its success, real-world usage often requires repeated inference to estimate statistical quantities for complex decision-making tasks involving uncertainty or extreme-value analysis, resulting in substantial latency. We introduce neural statistical functions, a new family of models learned from pre-trained single-sample predictors and scattered data samples, which can directly infer statistics over continuous operating condition ranges without explicit sampling. By introducing the notion of prefix statistics, we transform and unify diverse statistical functions (e.g., integrals, quantiles, and maxima) into an interval-conditional framework, in which a principled identity between the prefix statistics and the individual-case regression serves as the learning objective. Neural statistical functions achieve strong performance in estimating essential statistics of complex physical processes, including accumulated energy in dynamical systems, quantiles of aerodynamic responses, and maximum stress in crash processes, while achieving up to a 100$\times$ reduction in model evaluations.

cs.LG↗

arXiv2025

What's in the Box? Reasoning about Unseen Objects from Multimodal Cues

People regularly make inferences about objects in the world that they cannot see by flexibly integrating information from multiple sources: auditory and visual cues, language, and our prior beliefs and knowledge about the scene. How are we able to so flexibly integrate many sources of information to make sense of the world around us, even if we have no direct knowledge? In this work, we propose a neurosymbolic model that uses neural networks to parse open-ended multimodal inputs and then applies a Bayesian model to integrate different sources of information to evaluate different hypotheses. We evaluate our model with a novel object guessing game called ``What's in the Box?'' where humans and models watch a video clip of an experimenter shaking boxes and then try to guess the objects inside the boxes. Through a human experiment, we show that our model correlates strongly with human judgments, whereas unimodal ablated models and large multimodal neural model baselines show poor correlation.

cs.AI↗

arXiv2020

Modeling of Network Based Digital Contact Tracing and Testing Strategies for the COVID-19 Pandemic

With more than 1.7 million COVID-19 deaths, identifying effective measures to prevent COVID-19 is a top priority. We developed a mathematical model to simulate the COVID-19 pandemic with digital contact tracing and testing strategies. The model uses a real-world social network generated from a high-resolution contact data set of 180 students. This model incorporates infectivity variations, test sensitivities, incubation period, and asymptomatic cases. We present a method to extend the weighted temporal social network and present simulations on a network of 5000 students. The purpose of this work is to investigate optimal quarantine rules and testing strategies with digital contact tracing. The results show that the traditional strategy of quarantining direct contacts reduces infections by less than 20% without sufficient testing. Periodic testing every 2 weeks without contact tracing reduces infections by less than 3%. A variety of strategies are discussed including testing second and third degree contacts and the pre-exposure notification system, which acts as a social radar warning users how far they are from COVID-19. The most effective strategy discussed in this work was combined the pre-exposure notification system with testing second and third degree contacts. This strategy reduces infections by 18.3% when 30% of the population uses the app, 45.2% when 50% of the population uses the app, 72.1% when 70% of the population uses the app, and 86.8% when 95% of the population uses the app. When simulating the model on an extended network of 5000 students, the results are similar with the contact tracing app reducing infections by up to 79%.

cs.SI↗

arXiv2020

Proof of Kaneko--Tsumura Conjecture on Triple T-Values

Many $\mathbb{Q}$-linear relations exist between multiple zeta values, the most interesting of which are various weighted sum formulas. In this paper, we generalized these to Euler sums and some other variants of multiple zeta values by considering the generating functions of the Euler sums. Through this approach we are able to re-prove a few known formulas, confirm a conjecture of Kaneko and Tsumura on triple $T$-values, and discover many new identities.

math.NT↗

arXiv2019

Novel Non-Negative Variance Estimator for (Modified) Within-Cluster Resampling

This article proposes a novel variance estimator for within-cluster resampling (WCR) and modified within-cluster resampling (MWCR) - two existing methods for analyzing longitudinal data. WCR is a simple but computationally intensive method, in which a single observation is randomly sampled from each cluster to form a new dataset. This process is repeated numerous times, and in each resampled dataset (or outputation), we calculate beta using a generalized linear model. The final resulting estimator is an average across estimates from all outputations. MWCR is an extension of WCR that can account for the within-cluster correlation of the dataset; consequently, there are two noteworthy differences: 1) in MWCR, each resampled dataset is formed by randomly sampling multiple observations without replacement from each cluster and 2) generalized estimating equations (GEEs) are used to estimate the parameter of interest. While WCR and MWCR are relatively simple to implement, a key challenge is that the proposed moment-based estimator is often times negative in practice. Our modified variance estimator is not only strictly positive, but simulations show that it preserves the type I error and allows statistical power gains associated with MWCR to be realized.

stat.ME↗