PyTorchKR Papers

최신 논문

총 139,336편
태그·기간 등 상세 필터

카테고리 탐색

arXiv:2503.10699공식 코드 · Papers with Code

Test-Time Discovery via Hashing Memory

We introduce Test-Time Discovery (TTD) as a novel task that addresses class shifts during testing, requiring models to simultaneously identify emerging categories while preserving previously learned ones. A key challenge in TTD is distinguishing newly discovered classes from those already identified. To address this, we propose a training-free, hash-based memory mechanism that enhances class discovery through fine-grained comparisons with past test samples. Leveraging the characteristics of unknown classes, our approach introduces hash representation based on feature scale and directions, utilizing Locality-Sensitive Hashing (LSH) for efficient grouping of similar samples. This enables test samples to be easily and quickly compared with relevant past instances. Furthermore, we design a collaborative classification strategy, combining a prototype classifier for known classes with an LSH-based classifier for novel ones. To enhance reliability, we incorporate a self-correction mechanism that refines memory labels through hash-based neighbor retrieval, ensuring more stable and accurate class assignments. Experimental results demonstrate that our method achieves good discovery of novel categories while maintaining performance on known classes, establishing a new paradigm in model testing. Our code is available at https://github.com/fanlyu/ttd.

arXiv:2503.10687공식 코드 · Papers with Code

Context-guided Responsible Data Augmentation with Diffusion Models

Generative diffusion models offer a natural choice for data augmentation when training complex vision models. However, ensuring reliability of their generative content as augmentation samples remains an open challenge. Despite a number of techniques utilizing generative images to strengthen model training, it remains unclear how to utilize the combination of natural and generative images as a rich supervisory signal for effective model induction. In this regard, we propose a text-to-image (T2I) data augmentation method, named DiffCoRe-Mix, that computes a set of generative counterparts for a training sample with an explicitly constrained diffusion model that leverages sample-based context and negative prompting for a reliable augmentation sample generation. To preserve key semantic axes, we also filter out undesired generative samples in our augmentation process. To that end, we propose a hard-cosine filtration in the embedding space of CLIP. Our approach systematically mixes the natural and generative images at pixel and patch levels. We extensively evaluate our technique on ImageNet-1K,Tiny ImageNet-200, CIFAR-100, Flowers102, CUB-Birds, Stanford Cars, and Caltech datasets, demonstrating a notable increase in performance across the board, achieving up to $\sim 3\%$ absolute gain for top-1 accuracy over the state-of-the-art methods, while showing comparable computational overhead. Our code is publicly available at https://github.com/khawar-islam/DiffCoRe-Mix

arXiv:2503.10683공식 코드 · Papers with Code

Understanding the Quality-Diversity Trade-off in Diffusion Language Models

Diffusion models have seen immense success in modelling continuous data across a range of domains such as vision and audio. Despite the challenges of adapting diffusion models to discrete data, recent work explores their application to text generation by working in the continuous embedding space. However, these models lack a natural means to control the inherent trade-off between quality and diversity as afforded by the temperature hyperparameter in autoregressive models, hindering understanding of model performance and restricting generation quality. This work proposes the use of classifier-free guidance and stochastic clamping for manipulating the quality-diversity trade-off on sequence-to-sequence tasks, demonstrating that these techniques may be used to improve the performance of a diffusion language model.

arXiv:2503.10678공식 코드 · Papers with Code

VRMDiff: Text-Guided Video Referring Matting Generation of Diffusion

We propose a new task, video referring matting, which obtains the alpha matte of a specified instance by inputting a referring caption. We treat the dense prediction task of matting as video generation, leveraging the text-to-video alignment prior of video diffusion models to generate alpha mattes that are temporally coherent and closely related to the corresponding semantic instances. Moreover, we propose a new Latent-Constructive loss to further distinguish different instances, enabling more controllable interactive matting. Additionally, we introduce a large-scale video referring matting dataset with 10,000 videos. To the best of our knowledge, this is the first dataset that concurrently contains captions, videos, and instance-level alpha mattes. Extensive experiments demonstrate the effectiveness of our method. The dataset and code are available at https://github.com/Hansxsourse/VRMDiff.

arXiv:2503.01895공식 코드 · Papers with Code

Evaluating System 1 vs. 2 Reasoning Approaches for Zero-Shot Time Series Forecasting: A Benchmark and Insights

Reasoning ability is crucial for solving challenging tasks. With the advancement of foundation models, such as the emergence of large language models (LLMs), a wide range of reasoning strategies has been proposed, including test-time enhancements, such as Chain-ofThought, and post-training optimizations, as used in DeepSeek-R1. While these reasoning strategies have demonstrated effectiveness across various challenging language or vision tasks, their applicability and impact on time-series forecasting (TSF), particularly the challenging zero-shot TSF, remain largely unexplored. In particular, it is unclear whether zero-shot TSF benefits from reasoning and, if so, what types of reasoning strategies are most effective. To bridge this gap, we propose ReC4TS, the first benchmark that systematically evaluates the effectiveness of popular reasoning strategies when applied to zero-shot TSF tasks. ReC4TS conducts comprehensive evaluations across datasets spanning eight domains, covering both unimodal and multimodal with short-term and longterm forecasting tasks. More importantly, ReC4TS provides key insights: (1) Self-consistency emerges as the most effective test-time reasoning strategy; (2) Group-relative policy optimization emerges as a more suitable approach for incentivizing reasoning ability during post-training; (3) Multimodal TSF benefits more from reasoning strategies compared to unimodal TSF. Beyond these insights, ReC4TS establishes two pioneering starting blocks to support future zero-shot TSF reasoning research: (1) A novel dataset, TimeThinking, containing forecasting samples annotated with reasoning trajectories from multiple advanced LLMs, and (2) A new and simple test-time scaling-law validated on foundational TSF models enabled by self-consistency reasoning strategy. All data and code are publicly accessible at: https://github.com/AdityaLab/OpenTimeR

arXiv:2412.05334공식 코드 · Papers with Code

Closed-Loop Supervised Fine-Tuning of Tokenized Traffic Models

Traffic simulation aims to learn a policy for traffic agents that, when unrolled in closed-loop, faithfully recovers the joint distribution of trajectories observed in the real world. Inspired by large language models, tokenized multi-agent policies have recently become the state-of-the-art in traffic simulation. However, they are typically trained through open-loop behavior cloning, and thus suffer from covariate shift when executed in closed-loop during simulation. In this work, we present Closest Among Top-K (CAT-K) rollouts, a simple yet effective closed-loop fine-tuning strategy to mitigate covariate shift. CAT-K fine-tuning only requires existing trajectory data, without reinforcement learning or generative adversarial imitation. Concretely, CAT-K fine-tuning enables a small 7M-parameter tokenized traffic simulation policy to outperform a 102M-parameter model from the same model family, achieving the top spot on the Waymo Sim Agent Challenge leaderboard at the time of submission. The code is available at https://github.com/NVlabs/catk.

arXiv:2503.11532공식 코드 · Papers with Code

Observation-only learning of neural mapping schemes for gappy satellite-derived ocean colour parameters

Monitoring optical properties of coastal and open ocean waters is crucial to assessing the health of marine ecosystems. Deep learning offers a promising approach to address these ecosystem dynamics, especially in scenarios where gap-free ground-truth data is lacking, which poses a challenge for designing effective training frameworks. Using an advanced neural variational data assimilation scheme (called 4DVarNet), we introduce a comprehensive training framework designed to effectively train directly on gappy data sets. Using the Mediterranean Sea as a case study, our experiments not only highlight the high performance of the chosen neural network in reconstructing gap-free images from gappy datasets but also demonstrate its superior performance over state-of-the-art algorithms such as DInEOF and Direct Inversion, whether using CNN or UNet architectures.

arXiv:2503.11482공식 코드 · Papers with Code

NeuMC -- a package for neural sampling for lattice field theories

We present the \texttt{NeuMC} software package, based on \pytorch, aimed at facilitating the research on neural samplers in lattice field theories. Neural samplers based on normalizing flows are becoming increasingly popular in the context of Monte-Carlo simulations as they can effectively approximate target probability distributions, possibly alleviating some shortcomings of the Markov chain Monte-Carlo methods. Our package provides tools to create such samplers for two-dimensional field theories.

arXiv:2503.11470공식 코드 · Papers with Code

Topological Dictionary Learning

The aim of this paper is to introduce a novel dictionary learning algorithm for sparse representation of signals defined over combinatorial topological spaces, specifically, regular cell complexes. Leveraging Hodge theory, we embed topology into the dictionary structure via concatenated sub-dictionaries, each as a polynomial of Hodge Laplacians, yielding localized spectral topological filter frames. The learning problem is cast to jointly infer the underlying cell complex and optimize the dictionary coefficients and the sparse signal representation. We efficiently solve the problem via iterative alternating algorithms. Numerical results on both synthetic and real data show the effectiveness of the proposed procedure in jointly learning the sparse representations and the underlying relational structure of topological signals.

arXiv:2503.11427공식 코드 · Papers with Code

FlowKac: An Efficient Neural Fokker-Planck solver using Temporal Normalizing flows and the Feynman Kac-Formula

Solving the Fokker-Planck equation for high-dimensional complex dynamical systems remains a pivotal yet challenging task due to the intractability of analytical solutions and the limitations of traditional numerical methods. In this work, we present FlowKac, a novel approach that reformulates the Fokker-Planck equation using the Feynman-Kac formula, allowing to query the solution at a given point via the expected values of stochastic paths. A key innovation of FlowKac lies in its adaptive stochastic sampling scheme which significantly reduces the computational complexity while maintaining high accuracy. This sampling technique, coupled with a time-indexed normalizing flow, designed for capturing time-evolving probability densities, enables robust sampling of collocation points, resulting in a flexible and mesh-free solver. This formulation mitigates the curse of dimensionality and enhances computational efficiency and accuracy, which is particularly crucial for applications that inherently require dimensions beyond the conventional three. We validate the robustness and scalability of our method through various experiments on a range of stochastic differential equations, demonstrating significant improvements over existing techniques.

arXiv:2503.11359공식 코드 · Papers with Code

Learning-Based MPC for Fuel Efficient Control of Autonomous Vehicles with Discrete Gear Selection

Co-optimization of both vehicle speed and gear position via model predictive control (MPC) has been shown to offer benefits for fuel-efficient autonomous driving. However, optimizing both the vehicle's continuous dynamics and discrete gear positions may be too computationally intensive for a real-time implementation. This work proposes a learning-based MPC scheme to address this issue. A policy is trained to select and fix the gear positions across the prediction horizon of the MPC controller, leaving a significantly simpler continuous optimization problem to be solved online. In simulation, the proposed approach is shown to have a significantly lower computation burden and a comparable performance, with respect to pure MPC-based co-optimization.

arXiv:2503.11343공식 코드 · Papers with Code

FG-DFPN: Flow Guided Deformable Frame Prediction Network

Video frame prediction remains a fundamental challenge in computer vision with direct implications for autonomous systems, video compression, and media synthesis. We present FG-DFPN, a novel architecture that harnesses the synergy between optical flow estimation and deformable convolutions to model complex spatio-temporal dynamics. By guiding deformable sampling with motion cues, our approach addresses the limitations of fixed-kernel networks when handling diverse motion patterns. The multi-scale design enables FG-DFPN to simultaneously capture global scene transformations and local object movements with remarkable precision. Our experiments demonstrate that FG-DFPN achieves state-of-the-art performance on eight diverse MPEG test sequences, outperforming existing methods by 1dB PSNR while maintaining competitive inference speeds. The integration of motion cues with adaptive geometric transformations makes FG-DFPN a promising solution for next-generation video processing systems that require high-fidelity temporal predictions. The model and instructions to reproduce our results will be released at: https://github.com/KUIS-AI-Tekalp-Research Group/frame-prediction

arXiv:2503.11273공식 코드 · Papers with Code

Financial Fraud Detection with Entropy Computing

We introduce CVQBoost, a novel classification algorithm that leverages early hardware implementing Quantum Computing Inc's Entropy Quantum Computing (EQC) paradigm, Dirac-3 [Nguyen et. al. arXiv:2407.04512]. We apply CVQBoost to a fraud detection test case and benchmark its performance against XGBoost, a widely utilized ML method. Running on Dirac-3, CVQBoost demonstrates a significant runtime advantage over XGBoost, which we evaluate on high-performance hardware comprising up to 48 CPUs and four NVIDIA L4 GPUs using the RAPIDS AI framework. Our results show that CVQBoost maintains competitive accuracy (measured by AUC) while significantly reducing training time, particularly as dataset size and feature complexity increase. To assess scalability, we extend our study to large synthetic datasets ranging from 1M to 70M samples, demonstrating that CVQBoost on Dirac-3 is well-suited for large-scale classification tasks. These findings position CVQBoost as a promising alternative to gradient boosting methods, offering superior scalability and efficiency for high-dimensional ML applications such as fraud detection.

arXiv:2503.11213공식 코드 · Papers with Code

Simulating Dual-Pixel Images From Ray Tracing For Depth Estimation

Many studies utilize dual-pixel (DP) sensor phase characteristics for various applications, such as depth estimation and deblurring. However, since the DP image features are entirely determined by the camera hardware, DP-depth paired datasets are very scarce, especially when performing depth estimation on customized cameras. To overcome this, studies simulate DP images using ideal optical system models. However, these simulations often violate real optical propagation laws,leading to poor generalization to real DP data. To address this, we investigate the domain gap between simulated and real DP data, and propose solutions using the Simulating DP images from ray tracing (Sdirt) scheme. The Sdirt generates realistic DP images via ray tracing and integrates them into the depth estimation training pipeline. Experimental results show that models trained with Sdirt-simulated images generalize better to real DP data.

arXiv:2503.11123공식 코드 · Papers with Code

Modeling and Optimization for Flexible Cylindrical Arrays-Enabled Wireless Communications

Flexible-geometry arrays have garnered much attention in wireless communications, which dynamically adjust wireless channels to improve the system performance. In this paper, we propose a novel flexible-geometry array for a $360^\circ$ coverage, named flxible cylindrical array (FCLA), comprised of multiple flexible circular arrays (FCAs). The elements in each FCA can revolve around the circle track to change their horizontal positions, and the FCAs can move along the vertical axis to change the elements' heights. Considering that horizontal revolving can change the antenna orientation, we adopt both the omni-directional and the directional antenna patterns. Based on the regularized zero-forcing (RZF) precoding scheme, we formulate a particular compressive sensing (CS) problem incorporating joint precoding and antenna position optimization, and propose two effective methods, namely FCLA-J and FCLA-A, to solve it. Specifically, the first method involves jointly optimizing the element's revolving angle, height, and precoding coefficient within a single CS framework. The second method decouples the CS problem into two subproblems by utilizing an alternative sparse optimization approach for the revolving angle and height, thereby reducing time complexity. Simulation results reveal that, when utilizing directional radiation patterns, FCLA-J and FCLA-A achieve substantial performance improvements of 43.32\% and 25.42\%, respectively, compared to uniform cylindrical arrays (UCLAs) with RZF precoding.

arXiv:2503.11112공식 코드 · Papers with Code

Flexible Intelligent Metasurface-Aided Wireless Communications: Architecture and Performance

Typical reconfigurable intelligent surface (RIS) implementations include metasurfaces with almost passive unit elements capable of reflecting their incident waves in controllable ways, enhancing wireless communications in a cost-effective manner. In this paper, we advance the concept of intelligent metasurfaces by introducing a flexible array geometry, termed flexible intelligent metasurface (FIM), which supports both element movement (EM) and passive beamforming (PBF). In particular, based on the single-input single-output (SISO) system setup, we first compare three modes of FIM, namely, EM-only, PBF-only, and EM-PBF, in terms of received signal power under different FIM and channel setups. The PBF-only mode, which only adjusts the reflect phase, is shown to be less effective than the EM-only mode in enhancing received signal strength. In a multi-element, multi-path scenario, the EM-only mode improves the received signal power by 125% compared to the PBF-only mode. The EM-PBF mode, which optimizes both element positions and phases, further enhances performance. Additionally, we investigate the channel estimation problem for FIM systems by designing a protocol that gathers EM and PBF measurements, enabling the formulation of a compressive sensing problem for joint cascaded and direct channel estimation. We then propose a sparse recovery algorithm called clustering mean-field variational sparse Bayesian learning, which enhances estimation performance while maintaining low complexity.

arXiv:2503.11064공식 코드 · Papers with Code

MobiVital: Self-supervised Time-series Quality Estimation for Contactless Respiration Monitoring Using UWB Radar

Respiration waveforms are increasingly recognized as important biomarkers, offering insights beyond simple respiration rates, such as detecting breathing irregularities for disease diagnosis or monitoring breath patterns to guide rehabilitation training. Previous works in wireless respiration monitoring have primarily focused on estimating respiration rate, where the breath waveforms are often generated as a by-product. As a result, issues such as waveform deformation and inversion have largely been overlooked, reducing the signal's utility for applications requiring breathing waveforms. To address this problem, we present a novel approach, MobiVital, that improves the quality of respiration waveforms obtained from ultra-wideband (UWB) radar data. MobiVital combines a self-supervised autoregressive model for breathing waveform extraction with a biology-informed algorithm to detect and correct waveform inversions. To encourage reproducible research efforts for developing wireless vital signal monitoring systems, we also release a 12-person, 24-hour UWB radar vital signal dataset, with time-synchronized ground truth obtained from wearable sensors. Our results show that the respiration waveforms produced by our system exhibit a 7-34% increase in fidelity to the ground truth compared to the baselines and can benefit downstream tasks such as respiration rate estimation.

arXiv:2503.10950공식 코드 · Papers with Code

DNA Origami Nanostructures Observed in Transmission Electron Microscopy Images can be Characterized through Convolutional Neural Networks

Artificial intelligence (AI) models remain an emerging strategy to accelerate materials design and development. We demonstrate that convolutional neural network (CNN) models can characterize DNA origami nanostructures employed in programmable self-assembling, which is important in many applications such as in biomedicine. Specifically, we benchmark the performance of 9 CNN models -- viz. AlexNet, GoogLeNet, VGG16, VGG19, ResNet18, ResNet34, ResNet50, ResNet101, and ResNet152 -- to characterize the ligation number of DNA origami nanostructures in transmission electron microscopy (TEM) images. We first pre-train CNN models using a large image dataset of 720 images from our coarse-grained (CG) molecular dynamics (MD) simulations. Then, we fine-tune the pre-trained CNN models, using a small experimental TEM dataset with 146 TEM images. All CNN models were found to have similar computational time requirements, while their model sizes and performances are different. We use 20 test MD images to demonstrate that among all of the pre-trained CNN models ResNet50 and VGG16 have the highest and second highest accuracies. Among the fine-tuned models, VGG16 was found to have the highest agreement on the test TEM images. Thus, we conclude that fine-tuned VGG16 models can quickly characterize the ligation number of nanostructures in large TEM images.

arXiv:2503.10873공식 코드 · Papers with Code

Mamba time series forecasting with uncertainty quantification

State space models, such as Mamba, have recently garnered attention in time series forecasting due to their ability to capture sequence patterns. However, in electricity consumption benchmarks, Mamba forecasts exhibit a mean error of approximately 8\%. Similarly, in traffic occupancy benchmarks, the mean error reaches 18\%. This discrepancy leaves us to wonder whether the prediction is simply inaccurate or falls within error given spread in historical data. To address this limitation, we propose a method to quantify the predictive uncertainty of Mamba forecasts. Here, we propose a dual-network framework based on the Mamba architecture for probabilistic forecasting, where one network generates point forecasts while the other estimates predictive uncertainty by modeling variance. We abbreviate our tool, Mamba with probabilistic time series forecasting, as Mamba-ProbTSF and the code for its implementation is available on GitHub (https://github.com/PessoaP/Mamba-ProbTSF). Evaluating this approach on synthetic and real-world benchmark datasets, we find Kullback-Leibler divergence between the learned distributions and the data--which, in the limit of infinite data, should converge to zero if the model correctly captures the underlying probability distribution--reduced to the order of $10^{-3}$ for synthetic data and $10^{-1}$ for real-world benchmark, demonstrating its effectiveness. We find that in both the electricity consumption and traffic occupancy benchmark, the true trajectory stays within the predicted uncertainty interval at the two-sigma level about 95\% of the time. We end with a consideration of potential limitations, adjustments to improve performance, and considerations for applying this framework to processes for purely or largely stochastic dynamics where the stochastic changes accumulate, as observed for example in pure Brownian motion or molecular dynamics trajectories.

arXiv:2503.10865공식 코드 · Papers with Code

A theory of ecological invasions and its implications for eco-evolutionary dynamics

Predicting the outcomes of species invasions is a central goal of ecology, a task made especially challenging due to ecological feedbacks. To address this, we develop a general theory of ecological invasions applicable to a wide variety of ecological models: including Lotka-Volterra models, consumer resource models, and models with cross feeding. Importantly, our framework remains valid even when invading evolved (non-random) communities and accounts for invasion-driven species extinctions. We derive analytical expressions relating invasion fitness to invader abundance, shifts in the community, and extinction probabilities. These results can be understood through a new quantity we term ``dressed invasion fitness'', which augments the traditional notion of invasion fitness by incorporating ecological feedbacks. We apply our theory to analyze short-term evolutionary dynamics through a series of invasions by mutants whose traits are correlated with an existing parent. We demonstrate that, generically, mutants and parents can coexist, often by driving the extinction of low-abundance species. We validate theoretical predictions against experimental datasets spanning ecosystems from plants to microbial protists. Our work highlights the central role of ecological feedbacks in shaping community responses to invasions and mutations, suggesting that parent-mutant coexistence is widespread in eco-evolutionary dynamics.