Physical unclonable functions (PUFs) involve challenging practical applications of error-correcting codes (ECCs), requiring extremely low failure rates on the order of $10^{-6}$ and below despite raw input bit error rates as high as 22%. These requirements call for an efficient ultra-low rate code design. In this work, we propose a novel coding scheme tailored for PUFs based on Polar codes and a low-complexity version of automorphism ensemble decoding (AED). Notably, our serial AED scheme reuses a single successive cancellation (SC) decoder across multiple decoding attempts. By introducing cascaded and recursive interleavers, we efficiently scale the number of AED candidates without requiring expensive large multiplexers. An aggressive quantization strategy of only 3 bits per message further reduces the area requirements of the underlying SC decoder. The resulting coding scheme achieves the same block error rate of $10^{-6}$ as our baseline based on Bose-Ray-Chaudhuri-Hocquenghem (BCH) codes while requiring 1.75x fewer codeword bits to encode the same K = 312 payload bits. This reduction translates directly into 1.75x less helper data storage and, consequently, a smaller overall chip area.

Understanding and modeling lighting effects are fundamental tasks in computer vision and graphics. Classic physically-based rendering (PBR) accurately simulates the light transport, but relies on precise scene representations–explicit 3D geometry, high-quality material properties, and lighting conditions–that are often impractical to obtain in real-world scenarios. Therefore, we introduce DiffusionRenderer, a neural approach that addresses the dual problem of inverse and forward rendering within a holistic framework. Leveraging powerful video diffusion model priors, the inverse rendering model accurately estimates G-buffers from real-world videos, providing an interface for image editing tasks, and training data for the rendering model. Conversely, our rendering model generates photorealistic images from G-buffers without explicit light transport simulation. Specifically, we first train a video diffusion model for inverse rendering on synthetic data, which generalizes well to real-world videos and allows us to auto-label diverse real-world videos. We then co-train our rendering model using both synthetic and auto-labeled real-world data. Experiments demonstrate that DiffusionRenderer effectively approximates inverse and forwards rendering, consistently outperforming the state-of-the-art. Our model enables practical applications from a single video input—including relighting, material editing, and realistic object insertion.

We present Gen3C, a generative video model with precise Camera Control and temporal 3D Consistency. Prior video models already generate realistic videos, but they tend to leverage little 3D information, leading to inconsistencies, such as objects popping in and out of existence. Camera control, if implemented at all, is imprecise, because camera parameters are mere inputs to the neural network which must then infer how the video depends on the camera. In contrast, Gen3C is guided by a 3D cache: point clouds obtained by predicting the pixel-wise depth of seed images or previously generated frames. When generating the next frames, Gen3C is conditioned on the 2D renderings of the 3D cache with the new camera trajectory provided by the user. Crucially, this means that Gen3C neither has to remember what it previously generated nor does it have to infer the image structure from the camera pose. The model, instead, can focus all its generative power on previously unobserved regions, as well as advancing the scene state to the next frame. Our results demonstrate more precise camera control than prior work, as well as state-of-the-art results in sparse-view novel view synthesis, even in challenging settings such as driving scenes and monocular dynamic video. Results are best viewed in videos. Check out our webpage!1

Efficient deployment of small language models (SLMs) is essential for numerous real-world applications with stringent latency constraints.While previous work on SLM design has primarily focused on reducing the number of parameters to achieve parameter-optimal SLMs, parameter efficiency does not necessarily translate into proportional real-device speed-ups. This work aims to identify the key determinants of SLMs' real-device latency and offer generalizable principles and methodologies for SLM design and training when real-device latency is the primary consideration. Specifically, we identify two central architectural factors: depth–width ratios and operator choices. The former is crucial for small-batch-size latency, while the latter affects both latency and large-batch-size throughput. In light of this, we first study latency-optimal depth–width ratios, with the key finding that although deep–thin models generally achieve better accuracy under the same parameter budget, they may not lie on the accuracy–latency trade-off frontier. Next, we explore emerging efficient attention alternatives to evaluate their potential as candidate building operators. Using the identified promising operators, we construct an evolutionary search framework to automatically discover latency-optimal combinations of these operators within hybrid SLMs, thereby advancing the accuracy–latency frontier. In addition to architectural improvements, we further enhance SLM training using a weight normalization technique that enables more effective weight updates and improves final convergence. This technique can serve as a generalizable component for future SLMs. Combining these methods, we introduce a new family of hybrid SLMs, called Nemotron-Flash, which significantly advances the accuracy–efficiency frontier of state-of-the-art SLMs, e.g., achieving over +5.5\% average accuracy, 1.3$\times$/1.9$\times$ lower latency, and 18.7$\times$/45.6$\times$ higher throughput compared to Qwen3-1.7B/0.6B, respectively.

We address the challenge of relighting a single image or video, a task that demands precise scene intrinsic understanding and high-quality light transport synthesis. Existing end-to-end relighting models are often limited by the scarcity of paired multi-illumination data, restricting their ability to generalize across diverse scenes. Conversely, two-stage pipelines that combine inverse and forward rendering can mitigate data requirements but are susceptible to error accumulation and often fail to produce realistic outputs under complex lighting conditions or with sophisticated materials. In this work, we introduce a general-purpose approach that jointly estimates albedo and synthesizes relit outputs in a single pass, harnessing the generative capabilities of video diffusion models. This joint formulation enhances implicit scene comprehension and facilitates the creation of realistic lighting effects and intricate material interactions, such as shadows, reflections, and transparency. Trained on synthetic multi-illumination data and extensive automatically labeled real-world videos, our model demonstrates strong generalization across diverse domains and surpasses previous methods in both visual fidelity and temporal consistency. Our project page is https://research.nvidia.com/labs/toronto-ai/UniRelight/.

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 1.2–2.6 × 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.

We present a method for reducing errors---variance and bias---in physically based differentiable rendering (PBDR). Typical applications of PBDR repeatedly render a scene as part of an optimization loop involving gradient descent. The actual change introduced by each gradient descent step is often relatively small, causing a significant degree of redundancy in this computation. We exploit this redundancy by formulating a gradient estimator that employs a recursive control variate , which leverages information from previous optimization steps. The control variate reduces variance in gradients, and, perhaps more importantly, alleviates issues that arise from differentiating loss functions with respect to noisy inputs, a common cause of drift to bad local minima or divergent optimizations. We experimentally evaluate our approach on a variety of path-traced scenes containing surfaces and volumes and observe that primal rendering efficiency improves by a factor of up to 10.

Neural radiance fields achieve unprecedented quality for novel view synthesis, but their volumetric formulation remains expensive, requiring a huge number of samples to render high-resolution images. Volumetric encodings are essential to represent fuzzy geometry such as foliage and hair, and they are well-suited for stochastic optimization. Yet, many scenes ultimately consist largely of solid surfaces which can be accurately rendered by a single sample per pixel. Based on this insight, we propose a neural radiance formulation that smoothly transitions between volumetric- and surface-based rendering, greatly accelerating rendering speed and even improving visual fidelity. Our method constructs an explicit mesh envelope which spatially bounds a neural volumetric representation. In solid regions, the envelope nearly converges to a surface and can often be rendered with a single sample. To this end, we generalize the NeuS [Wang et al. 2021] formulation with a learned spatially-varying kernel size which encodes the spread of the density, fitting a wide kernel to volume-like regions and a tight kernel to surface-like regions. We then extract an explicit mesh of a narrow band around the surface, with width determined by the kernel size, and fine-tune the radiance field within this band. At inference time, we cast rays against the mesh and evaluate the radiance field only within the enclosed region, greatly reducing the number of samples required. Experiments show that our approach enables efficient rendering at very high fidelity. We also demonstrate that the extracted envelope enables downstream applications such as animation and simulation.

We propose a neural network (NN)-based algorithm for device detection and time of arrival (ToA) and carrier frequency offset (CFO) estimation for the narrowband physical random-access channel (NPRACH) of narrowband internet of things (NB-IoT). The introduced NN architecture leverages residual convolutional networks as well as knowledge of the preamble structure of the 5G New Radio (5G NR) specifications. Benchmarking on a 3rd Generation Partnership Project (3GPP) urban microcell (UMi) channel model with random drops of users against a state-of-the-art baseline shows that the proposed method enables up to 8 dB gains in false negative rate (FNR) as well as significant gains in false positive rate (FPR) and ToA and CFO estimation accuracy. Moreover, our simulations indicate that the proposed algorithm enables gains over a wide range of channel conditions, CFOs, and transmission probabilities. The introduced synchronization method operates at the base station (BS) and, therefore, introduces no additional complexity on the user devices. It could lead to an extension of battery lifetime by reducing the preamble length or the transmit power. Our code is available at: https://github.com/NVlabs/nprach_synch/.

Abstract Interactive global illumination remains a challenge in radiometrically and geometrically complex scenes. Specialized sampling strategies are effective for specular and near‐specular transport because the scattering has relatively low directional variance per scattering event. In contrast, the high variance from transport paths comprising multiple rough glossy or diffuse scattering events remains notoriously difficult to resolve with a small number of samples. We extend unidirectional path tracing to address this by combining screen‐space reservoir resampling and sparse world‐space probes, significantly improving sample efficiency for transport contributions that terminate on diffuse scattering events. Our experiments demonstrate a clear improvement—at equal time and equal quality—over purely path traced and purely probe‐based baselines. Moreover, when combined with commodity denoisers, we are able to interactively render global illumination in complex scenes.

Neural graphics primitives, parameterized by fully connected neural networks, can be costly to train and evaluate. We reduce this cost with a versatile new input encoding that permits the use of a smaller network without sacrificing quality, thus significantly reducing the number of floating point and memory access operations: a small neural network is augmented by a multiresolution hash table of trainable feature vectors whose values are optimized through stochastic gradient descent. The multiresolution structure allows the network to disambiguate hash collisions, making for a simple architecture that is trivial to parallelize on modern GPUs. We leverage this parallelism by implementing the whole system using fully-fused CUDA kernels with a focus on minimizing wasted bandwidth and compute operations. We achieve a combined speedup of several orders of magnitude, enabling training of high-quality neural graphics primitives in a matter of seconds, and rendering in tens of milliseconds at a resolution of 1920×1080.

Volumetric representations are popular in inverse rendering because they have a simple parameterization, are smoothly varying, and transparently handle topology changes. However, incorporating the full volumetric transport of light is costly and challenging, often leading practitioners to implement simplified models, such as purely emissive and absorbing volumes with "baked" lighting. One such challenge is the efficient estimation of the gradients of the volume's appearance with respect to its scattering and absorption parameters. We show that the straightforward approach---differentiating a volumetric free-flight sampler---can lead to biased and high-variance gradients, hindering optimization. Instead, we propose using a new sampling strategy: differential ratio tracking , which is unbiased, yields low-variance gradients, and runs in linear time. Differential ratio tracking combines ratio tracking and reservoir sampling to estimate gradients by sampling distances proportional to the unweighted transmittance rather than the usual extinction-weighted transmittance. In addition, we observe local minima when optimizing scattering parameters to reproduce dense volumes or surfaces. We show that these local minima can be overcome by bootstrapping the optimization from nonphysical emissive volumes that are easily optimized.

Many applications ranging from quasi-Monte Carlo integration over optimal control to neural networks benefit from high-dimensional, highly uniform samples. In the case of computer graphics, and more particularly in rendering, despite the need for uniformity, several sub-problems expose a low-dimensional structure. In this context, mastering sampling uniformity over projections while preserving high-dimensional uniformity has been intrinsically challenging. This difficulty may explain the relatively small number of mathematical constructions for such samplers. We propose a novel approach by showing that uniformity constraints can be expressed as an integer linear program that results in a sampler with the desired properties. As it turns out, complex constraints are easy to describe by means of stratification and sequence properties of digital nets. Formalized using generator matrix determinants, our new MatBuilder software solves the set of constraints by iterating the linear integer program solver in a greedy fashion to compute a problem-specific set of generator matrices that can be used as a drop-in replacement in the popular digital net samplers. The samplers created by MatBuilder achieve the uniformity of classic low discrepancy sequences. More importantly, we demonstrate the benefit of the unprecedented versatility of our constraint approach with respect to low-dimensional problem structure for several applications.

We demonstrate that 1x1-convolutions in 1D time-channel separable convolutions may be replaced by constant, sparse random ternary matrices with weights in $\{-1,0,+1\}$. Such layers do not perform any multiplications and do not require training. Moreover, the matrices may be generated on the chip during computation and therefore do not require any memory access. With the same parameter budget, we can afford deeper and more expressive models, improving the Pareto frontiers of existing models on several tasks. For command recognition on Google Speech Commands v1, we improve the state-of-the-art accuracy from $97.21\%$ to $97.41\%$ at the same network size. Alternatively, we can lower the cost of existing models. For speech recognition on Librispeech, we half the number of weights to be trained while only sacrificing about $1\%$ of the floating-point baseline's word error rate.

We present a real-time neural radiance caching method for path-traced global illumination. Our system is designed to handle fully dynamic scenes, and makes no assumptions about the lighting, geometry, and materials. The data-driven nature of our approach sidesteps many difficulties of caching algorithms, such as locating, interpolating, and updating cache points. Since pretraining neural networks to handle novel, dynamic scenes is a formidable generalization challenge, we do away with pretraining and instead achieve generalization via adaptation , i.e. we opt for training the radiance cache while rendering. We employ self-training to provide low-noise training targets and simulate infinite-bounce transport by merely iterating few-bounce training updates. The updates and cache queries incur a mild overhead---about 2.6ms on full HD resolution---thanks to a streaming implementation of the neural network that fully exploits modern hardware. We demonstrate significant noise reduction at the cost of little induced bias, and report state-of-the-art, real-time performance on a number of challenging scenarios.

Rendering quality is largely influenced by the samplers used in Monte Carlo integration. Important factors include sample uniformity (e.g., low discrepancy) in the high-dimensional integration domain, sample uniformity in lower-dimensional projections, and lack of dominant structures that could result in aliasing artifacts. A widely used and successful construction is the Sobol' sequence that guarantees good high-dimensional uniformity and consequently results in faster convergence of quasi-Monte Carlo integration. We show that this sequence exhibits low uniformity and dominant structures in low-dimensional projections. These structures impair quality in the context of rendering, as they precisely occur in the 2-dimensional projections used for sampling light sources, reflectance functions, or the camera lens or sensor. We propose a new cascaded construction, which, despite dropping the sequential aspect of Sobol' samples, produces point sets exhibiting provably perfect dyadic partitioning (and therefore, excellent uniformity) in consecutive 2-dimensional projections, while preserving good high-dimensional uniformity. By optimizing the initialization parameters and performing Owen scrambling at finer levels of binary representations, we further improve over Sobol's integration convergence rate. Our method does not incur any overhead as compared to the generation of the Sobol' sequence, is compatible with Owen scrambling and can be used in rendering applications.

We propose neural control variates (NCV) for unbiased variance reduction in parametric Monte Carlo integration. So far, the core challenge of applying the method of control variates has been finding a good approximation of the integrand that is cheap to integrate. We show that a set of neural networks can face that challenge: a normalizing flow that approximates the shape of the integrand and another neural network that infers the solution of the integral equation. We also propose to leverage a neural importance sampler to estimate the difference between the original integrand and the learned control variate. To optimize the resulting parametric estimator, we derive a theoretically optimal, variance-minimizing loss function, and propose an alternative, composite loss for stable online training in practice. When applied to light transport simulation, neural control variates are capable of matching the state-of-the-art performance of other unbiased approaches, while providing means to develop more performant, practical solutions. Specifically, we show that the learned light-field approximation is of sufficient quality for high-order bounces, allowing us to omit the error correction and thereby dramatically reduce the noise at the cost of negligible visible bias.

We describe a decomposition for in-place matrix transposition, with applications to Array of Structures memory accesses on SIMD processors. Traditional approaches to in-place matrix transposition involve cycle following, which is difficult to parallelize, and on matrices of dimension m by n require O(mn log mn) work when limited to less than O(mn) auxiliary space. Our decomposition allows the rows and columns to be operated on independently during in-place transposition, reducing work complexity to O(mn), given O(max(m, n)) auxiliary space. This decomposition leads to an efficient and naturally parallel algorithm: we have measured median throughput of 19.5 GB/s on an NVIDIA Tesla K20c processor. An implementation specialized for the skinny matrices that arise when converting Arrays of Structures to Structures of Arrays yields median throughput of 34.3 GB/s, and a maximum throughput of 51 GB/s.

Abstract We present a physically based progressive global illumination system that is capable of simulating complex lighting situations robustly by efficiently using both light and eye paths. Specifically, we combine three distinct algorithms: point‐light‐based illumination which produces low‐noise approximations for diffuse inter‐reflections, specular gathering for glossy and singular effects and a caustic histogram method for the remaining light paths. The combined system efficiently renders low‐noise production quality images with indirect illumination from arbitrary light sources including inter‐reflections from caustics and allows for simulating depth of field and dispersion effects. Our system computes progressive approximations by continuously refining the solution using a constant memory footprint without the need of pre‐computations or optimizing parameters beforehand.

Abstract Storing textures on orthogonal tensor product lattices is predominant in computer graphics, although it is known that their sampling efficiency is not optimal. In two dimensions, the hexagonal lattice provides the maximum sampling efficiency. However, handling these lattices is difficult, because they are not able to tile an arbitrary rectangular region and have an irrational basis. By storing textures on rank‐1 lattices, we resolve both problems: Rank‐1 lattices can closely approximate hexagonal lattices, while all coordinates of the lattice points remain integer. At identical memory footprint texture quality is improved as compared to traditional orthogonal tensor product lattices due to the higher sampling efficiency. We introduce the basic theory of rank‐1 lattice textures and present an algorithmic framework which easily can be integrated into existing off‐line and real‐time rendering systems.

Abstract Photorealistic image synthesis is a computationally demanding task that relies on ray tracing for the evaluation of integrals. Rendering time is dominated by tracing long paths that are very incoherent by construction. We therefore investigate the use of SIMD instructions to accelerate incoherent rays. SIMD is used in the hierarchy construction, the tree traversal and the leaf intersection. This is achieved by increasing the arity of acceleration structures, which also reduces memory requirements. We show that the resulting hierarchies can be built quickly and are smaller than acceleration structures known so far while at the same time outperforming them for incoherent rays. Our new acceleration structure speeds up ray tracing by a factor of 1.6 to 2.0 compared to a highly optimized bounding interval hierarchy implementation, and 1.3 to 1.6 compared to an efficient kd‐tree. At the same time, the memory requirements are reduced by 10–50%. Additionally we show how a caching mechanism in conjunction with this memory efficient hierarchy can be used to speed up shadow rays in a global illumination algorithm without increasing the memory footprint. This optimization decreased the number of traversal steps up to 50%.

Image synthesis often requires the Monte Carlo estimation of integrals. Based on a generalized concept of stratification we present an efficient sampling scheme that consistently outperforms previous techniques. This is achieved by assembling sampling patterns that are stratified in the sense of jittered sampling and N-rooks sampling at the same time. The faster convergence and improved anti-aliasing are demonstrated by numerical experiments. Categories and Subject Descriptors (according to ACM CCS): G.3 [Probability and Statistics]: Probabilistic Algorithms (including Monte Carlo); I.3.2 [Computer Graphics]: Picture/Image Generation; I.3.7 [Computer Graphics]: Three-Dimensional Graphics and Realism.

We present a fundamental procedure for instant rendering from the radiance equation.Operating directly on the textured scene description, the very efficient and simple algorithm produces photorealistic images without any finite element kernel or solution discretization of the underlying integral equation.Rendering rates of a few seconds are obtained by exploiting graphics hardware, the deterministic technique of the quasi-random walk for the solution of the global illumination problem, and the new method of jittered low discrepancy sampling.