arXiv subjects

Alec Jacobson

Publications and source records attributed to Alec Jacobson.

At least 19 recordsLinked to original sources

arXiv2026

Remember to be Curious: Episodic Context and Persistent Worlds for 3D Exploration

Exploration is a prerequisite for learning useful behaviors in sparse-reward, long-horizon tasks, particularly within 3D environments. Curiosity-driven reinforcement learning addresses this via intrinsic rewards derived from the mismatch between the agent's predictive model of the world and reality. However, translating this intrinsic motivation to complex, photorealistic environments remains difficult, as agents can become trapped in local loops and receive fresh rewards for revisiting forgotten states. In this work, we demonstrate that this failure stems from a lack of spatial persistence and episodic context. We show that effective curiosity requires a model of the world that is persistent and continuously updated, paired with an agent that maintains an episodic trajectory history to navigate toward novel regions. We achieve this using an online 3D reconstruction as a persistent model of the world, while the agent policy is parameterized as a sequence model over RGB observations to maintain episodic context. This design enables effective exploration during training while allowing the agent to navigate using solely RGB frames at deployment. Trained purely via curiosity on HM3D, our agent outperforms RL-based active mapping baselines and generalizes zero-shot to Gibson and AI-generated worlds. Our end-to-end policy enables efficient adaptation to downstream tasks, such as apple picking and image-goal navigation, outperforming from-scratch baselines. Please see video results at https://recuriosity.github.io/.

cs.LG↗

arXiv2026

Progressive Convex Hull Simplification

Convex hulls are useful as tight bounding proxies for a variety of tasks including collision detection, ray intersection, and distance computation. Unfortunately, the complexity of polyhedral convex hulls grows linearly with their input. We consider the problem of conservatively simplifying a convex hull to a specified number of half-spaces while minimizing added volume or surface area. By working in the dual representation, we propose an efficient $O(n \log n)$ greedy optimization. In comparisons, we show that existing methods either exhibit poor efficiency, tightness or safety. We demonstrate the success of our method on a variety of input shapes and downstream application domains.

cs.GR↗

arXiv2026

GimmBO: Interactive Generative Image Model Merging via Bayesian Optimization

Fine-tuning-based adaptation is widely used to customize diffusion-based image generation, leading to large collections of community-created adapters that capture diverse subjects and styles. Adapters derived from the same base model can be merged with weights, enabling the synthesis of new visual results within a vast and continuous design space. To explore this space, current workflows rely on manual slider-based tuning, an approach that scales poorly and makes weight selection difficult, even when the candidate set is limited to 20-30 adapters. We propose GimmBO to support interactive exploration of adapter merging for image generation through Preferential Bayesian Optimization (PBO). Motivated by observations from real-world usage, including sparsity and constrained weight ranges, we introduce a two-stage BO backend that improves sampling efficiency and convergence in high-dimensional spaces. We evaluate our approach with simulated users and a user study, demonstrating improved convergence, high success rates, and consistent gains over BO and line-search baselines, and further show the flexibility of the framework through several extensions.

cs.CV↗

arXiv2025

Differentiable Variable Fonts

Editing and animating text appearance for graphic designs, commercials, etc. remain highly skilled tasks requiring detailed, hands on efforts from artists. Automating these manual workflows requires balancing the competing goals of maintaining legibility and aesthetics of text, while enabling creative expression. Variable fonts, recent parametric extensions to traditional fonts, offer the promise of new ways to ease and automate typographic design and animation. Variable fonts provide custom constructed parameters along which fonts can be smoothly varied. These parameterizations could then potentially serve as high value continuous design spaces, opening the door to automated design optimization tools. However, currently variable fonts are underutilized in creative applications, because artists so far still need to manually tune font parameters. Our work opens the door to intuitive and automated font design and animation workflows with differentiable variable fonts. To do so we distill the current variable font specification to a compact mathematical formulation that differentiably connects the highly non linear, non invertible mapping of variable font parameters to the underlying vector graphics representing the text. This enables us to construct a differentiable framework, with respect to variable font parameters, allowing us to perform gradient based optimization of energies defined on vector graphics control points, and on target rasterized images. We demonstrate the utility of this framework with four applications: direct shape manipulation, overlap aware modeling, physics based text animation, and automated font design optimization. Our work now enables leveraging the carefully designed affordances of variable fonts with differentiability to use modern design optimization technologies, opening new possibilities for easy and intuitive typographic design workflows.

cs.GR↗

arXiv2025

Mesh Processing Non-Meshes via Neural Displacement Fields

Mesh processing pipelines are mature, but adapting them to newer non-mesh surface representations -- which enable fast rendering with compact file size -- requires costly meshing or transmitting bulky meshes, negating their core benefits for streaming applications. We present a compact neural field that enables common geometry processing tasks across diverse surface representations. Given an input surface, our method learns a neural map from its coarse mesh approximation to the surface. The full representation totals only a few hundred kilobytes, making it ideal for lightweight transmission. Our method enables fast extraction of manifold and Delaunay meshes for intrinsic shape analysis, and compresses scalar fields for efficient delivery of costly precomputed results. Experiments and applications show that our fast, compact, and accurate approach opens up new possibilities for interactive geometry processing.

cs.GR↗

arXiv2025

Uniform Sampling of Surfaces by Casting Rays

Randomly sampling points on surfaces is an essential operation in geometry processing. This sampling is computationally straightforward on explicit meshes, but it is much more difficult on other shape representations, such as widely-used implicit surfaces. This work studies a simple and general scheme for sampling points on a surface, which is derived from a connection to the intersections of random rays with the surface. Concretely, given a subroutine to cast a ray against a surface and find all intersections, we can use that subroutine to uniformly sample white noise points on the surface. This approach is particularly effective in the context of implicit signed distance functions, where sphere marching allows us to efficiently cast rays and sample points, without needing to extract an intermediate mesh. We analyze the basic method to show that it guarantees uniformity, and find experimentally that it is significantly more efficient than alternative strategies on a variety of representations. Furthermore, we show extensions to blue noise sampling and stratified sampling, and applications to deform neural implicit surfaces as well as moment estimation.

cs.GR↗

arXiv2025

Stochastic Preconditioning for Neural Field Optimization

Neural fields are a highly effective representation across visual computing. This work observes that fitting these fields is greatly improved by incorporating spatial stochasticity during training, and that this simple technique can replace or even outperform custom-designed hierarchies and frequency space constructions. The approach is formalized as implicitly operating on a blurred version of the field, evaluated in-expectation by sampling with Gaussian-distributed offsets. Querying the blurred field during optimization greatly improves convergence and robustness, akin to the role of preconditioners in numerical linear algebra. This implicit, sampling-based perspective fits naturally into the neural field paradigm, comes at no additional cost, and is extremely simple to implement. We describe the basic theory of this technique, including details such as handling boundary conditions, and extending to a spatially-varying blur. Experiments demonstrate this approach on representations including coordinate MLPs, neural hashgrids, triplanes, and more, across tasks including surface reconstruction and radiance fields. In settings where custom-designed hierarchies have already been developed, stochastic preconditioning nearly matches or improves their performance with a simple and unified approach; in settings without existing hierarchies it provides an immediate boost to quality and robustness.

cs.GR↗

arXiv2025

Rigid Body Adversarial Attacks

Due to their performance and simplicity, rigid body simulators are often used in applications where the objects of interest can considered very stiff. However, no material has infinite stiffness, which means there are potentially cases where the non-zero compliance of the seemingly rigid object can cause a significant difference between its trajectories when simulated in a rigid body or deformable simulator. Similarly to how adversarial attacks are developed against image classifiers, we propose an adversarial attack against rigid body simulators. In this adversarial attack, we solve an optimization problem to construct perceptually rigid adversarial objects that have the same collision geometry and moments of mass to a reference object, so that they behave identically in rigid body simulations but maximally different in more accurate deformable simulations. We demonstrate the validity of our method by comparing simulations of several examples in commercially available simulators.

cs.CV↗

arXiv2024

A LoRA is Worth a Thousand Pictures

Recent advances in diffusion models and parameter-efficient fine-tuning (PEFT) have made text-to-image generation and customization widely accessible, with Low Rank Adaptation (LoRA) able to replicate an artist's style or subject using minimal data and computation. In this paper, we examine the relationship between LoRA weights and artistic styles, demonstrating that LoRA weights alone can serve as an effective descriptor of style, without the need for additional image generation or knowledge of the original training set. Our findings show that LoRA weights yield better performance in clustering of artistic styles compared to traditional pre-trained features, such as CLIP and DINO, with strong structural similarities between LoRA-based and conventional image-based embeddings observed both qualitatively and quantitatively. We identify various retrieval scenarios for the growing collection of customized models and show that our approach enables more accurate retrieval in real-world settings where knowledge of the training images is unavailable and additional generation is required. We conclude with a discussion on potential future applications, such as zero-shot LoRA fine-tuning and model attribution.

cs.CV↗

arXiv2024

RoMo: Robust Motion Segmentation Improves Structure from Motion

There has been extensive progress in the reconstruction and generation of 4D scenes from monocular casually-captured video. While these tasks rely heavily on known camera poses, the problem of finding such poses using structure-from-motion (SfM) often depends on robustly separating static from dynamic parts of a video. The lack of a robust solution to this problem limits the performance of SfM camera-calibration pipelines. We propose a novel approach to video-based motion segmentation to identify the components of a scene that are moving w.r.t. a fixed world frame. Our simple but effective iterative method, RoMo, combines optical flow and epipolar cues with a pre-trained video segmentation model. It outperforms unsupervised baselines for motion segmentation as well as supervised baselines trained from synthetic data. More importantly, the combination of an off-the-shelf SfM pipeline with our segmentation masks establishes a new state-of-the-art on camera calibration for scenes with dynamic content, outperforming existing methods by a substantial margin.

cs.CV↗

arXiv2024

Trust-Region Eigenvalue Filtering for Projected Newton

We introduce a novel adaptive eigenvalue filtering strategy to stabilize and accelerate the optimization of Neo-Hookean energy and its variants under the Projected Newton framework. For the first time, we show that Newton's method, Projected Newton with eigenvalue clamping and Projected Newton with absolute eigenvalue filtering can be unified using ideas from the generalized trust region method. Based on the trust-region fit, our model adaptively chooses the correct eigenvalue filtering strategy to apply during the optimization. Our method is simple but effective, requiring only two lines of code change in the existing Projected Newton framework. We validate our model outperforms stand-alone variants across a number of experiments on quasistatic simulation of deformable solids over a large dataset.

cs.GR↗

arXiv2024

2D Neural Fields with Learned Discontinuities

Effective representation of 2D images is fundamental in digital image processing, where traditional methods like raster and vector graphics struggle with sharpness and textural complexity respectively. Current neural fields offer high-fidelity and resolution independence but require predefined meshes with known discontinuities, restricting their utility. We observe that by treating all mesh edges as potential discontinuities, we can represent the magnitude of discontinuities with continuous variables and optimize. Based on this observation, we introduce a novel discontinuous neural field model that jointly approximate the target image and recovers discontinuities. Through systematic evaluations, our neural field demonstrates superior performance in denoising and super-resolution tasks compared to InstantNGP, achieving improvements of over 5dB and 10dB, respectively. Our model also outperforms Mumford-Shah-based methods in accurately capturing discontinuities, with Chamfer distances 3.5x closer to the ground truth. Additionally, our approach shows remarkable capability in handling complex artistic drawings and natural images.

cs.CV↗

arXiv2024

SpotlessSplats: Ignoring Distractors in 3D Gaussian Splatting

3D Gaussian Splatting (3DGS) is a promising technique for 3D reconstruction, offering efficient training and rendering speeds, making it suitable for real-time applications.However, current methods require highly controlled environments (no moving people or wind-blown elements, and consistent lighting) to meet the inter-view consistency assumption of 3DGS. This makes reconstruction of real-world captures problematic. We present SpotLessSplats, an approach that leverages pre-trained and general-purpose features coupled with robust optimization to effectively ignore transient distractors. Our method achieves state-of-the-art reconstruction quality both visually and quantitatively, on casual captures. Additional results available at: https://spotlesssplats.github.io

cs.CV↗

arXiv2024

Cascading upper bounds for triangle soup Pompeiu-Hausdorff distance

We propose a new method to accurately approximate the Pompeiu-Hausdorff distance from a triangle soup A to another triangle soup B up to a given tolerance. Based on lower and upper bound computations, we discard triangles from A that do not contain the maximizer of the distance to B and subdivide the others for further processing. In contrast to previous methods, we use four upper bounds instead of only one, three of which newly proposed by us. Many triangles are discarded using the simpler bounds, while the most difficult cases are dealt with by the other bounds. Exhaustive testing determines the best ordering of the four upper bounds. A collection of experiments shows that our method is faster than all previous accurate methods in the literature.

cs.GR↗

arXiv2024

Optimized Dual-Volumes for Tetrahedral Meshes

Constructing well-behaved Laplacian and mass matrices is essential for tetrahedral mesh processing. Unfortunately, the \emph{de facto} standard linear finite elements exhibit bias on tetrahedralized regular grids, motivating the development of finite-volume methods. In this paper, we place existing methods into a common construction, showing how their differences amount to the choice of simplex centers. These choices lead to satisfaction or breakdown of important properties: continuity with respect to vertex positions, positive semi-definiteness of the implied Dirichlet energy, positivity of the mass matrix, and unbiased-ness on regular grids. Based on this analysis, we propose a new method for constructing dual-volumes which explicitly satisfy all of these properties via convex optimization.

cs.GR↗

arXiv2024

Stabler Neo-Hookean Simulation: Absolute Eigenvalue Filtering for Projected Newton

Volume-preserving hyperelastic materials are widely used to model near-incompressible materials such as rubber and soft tissues. However, the numerical simulation of volume-preserving hyperelastic materials is notoriously challenging within this regime due to the non-convexity of the energy function. In this work, we identify the pitfalls of the popular eigenvalue clamping strategy for projecting Hessian matrices to positive semi-definiteness during Newton's method. We introduce a novel eigenvalue filtering strategy for projected Newton's method to stabilize the optimization of Neo-Hookean energy and other volume-preserving variants under high Poisson's ratio (near 0.5) and large initial volume change. Our method only requires a single line of code change in the existing projected Newton framework, while achieving significant improvement in both stability and convergence speed. We demonstrate the effectiveness and efficiency of our eigenvalue projection scheme on a variety of challenging examples and over different deformations on a large dataset.

cs.GR↗

arXiv2024

Text-guided Controllable Mesh Refinement for Interactive 3D Modeling

We propose a novel technique for adding geometric details to an input coarse 3D mesh guided by a text prompt. Our method is composed of three stages. First, we generate a single-view RGB image conditioned on the input coarse geometry and the input text prompt. This single-view image generation step allows the user to pre-visualize the result and offers stronger conditioning for subsequent multi-view generation. Second, we use our novel multi-view normal generation architecture to jointly generate six different views of the normal images. The joint view generation reduces inconsistencies and leads to sharper details. Third, we optimize our mesh with respect to all views and generate a fine, detailed geometry as output. The resulting method produces an output within seconds and offers explicit user control over the coarse structure, pose, and desired details of the resulting 3D mesh.

cs.CV↗

arXiv2023

Compact Neural Graphics Primitives with Learned Hash Probing

Neural graphics primitives are faster and achieve higher quality when their neural networks are augmented by spatial data structures that hold trainable features arranged in a grid. However, existing feature grids either come with a large memory footprint (dense or factorized grids, trees, and hash tables) or slow performance (index learning and vector quantization). In this paper, we show that a hash table with learned probes has neither disadvantage, resulting in a favorable combination of size and speed. Inference is faster than unprobed hash tables at equal quality while training is only 1.2-2.6x slower, significantly outperforming prior index learning approaches. We arrive at this formulation by casting all feature grids into a common framework: they each correspond to a lookup function that indexes into a table of feature vectors. In this framework, the lookup functions of existing data structures can be combined by simple arithmetic combinations of their indices, resulting in Pareto optimal compression and speed.

cs.CV↗