#generative modeling
31 papers match
EgoGVAE: Ego-body Mesh Reconstruction via Guided Variational Autoencoder
Jaehun Jung, Wonjun Kim
The paper proposes a guided variational autoencoder that reconstructs a full-body mesh from only head pose information, achieving fast one-step inference that is over 50 times fast…
Cosmo-SPINN: Fuzzy Dark Matter Simulations with Physics-Informed Generative Networks
Ashutosh Kumar Mishra, Emma Tolley, Nicolas Cerardi
The paper introduces a physics‑informed generative U‑Net that can evolve fuzzy dark matter fields and perform super‑resolution of simulations while enforcing the Schrödinger‑Poisso…
SE(3)-MeanFlow: Few-Step Protein Backbone Generation on Lie Groups
Yikun Bai, Binghang Lu, Yikai Liu +7
The paper presents SE(3)-MeanFlow, a generative model that creates protein backbone structures directly on the SE(3) Lie group using a few inference steps, avoiding costly ODE inte…
Existence-Field Diffusion Model for Spatial Point Processes with Variable Cardinality
Xiaoyin Pan, Christian R. Shelton, Rakshith Mahishi +1
The paper introduces the existence-field diffusion model, which uses an existence variable for each potential point to jointly model spatial locations and the number of points in s…
Flow Map Learning via Nongradient Vector Flow
Mark Goldstein, Anshuk Uppal, Raghav Singhal +2
The paper proposes SGFlow, a method that learns flow maps for diffusion and flow‑based generative models without requiring model invertibility or backpropagation through repeated m…
Equilibrium Training of Energy-Based Models with Parallel Trajectory Tempering
Nicolas Béreux, Aurélien Decelle, Cyril Furtlehner +1
The paper proposes Parallel Trajectory Tempering (PTT), a training method that keeps equilibrium sampling throughout learning of energy‑based models, enabling fast and stable train…
Galvanic Vestibular Stimulation in Latent Space
Zhi Liu, Tatsuki Fushimi, Yoichi Ochiai
The paper presents a dataset linking galvanic vestibular stimulation (GVS) waveforms to free-form sensation descriptions and introduces a retrieval‑guided variational autoencoder t…
Explorative Modeling: Unlocking a Third Pretraining Axis and End-to-End Generation
Alexi Gladstone, Heng Ji, Yilun Du
The paper proposes Explorative Modeling, a new training paradigm that selects the best among multiple candidate generations to improve generative models, adding a third pretraining…
MIDI-RAE-JEPA: Hierarchical Representation Learning and Generation for Symbolic Music
Scott H. Hawley
The paper introduces MIDI-RAE-JEPA, a self‑supervised model that learns hierarchical, equivariant representations of symbolic music from piano‑roll images using a Swin Transformer…
Trajectory-Aware Flow Matching for Topology Optimisation
Shusheng Xiao, Jinshuai Bai, Hyogu Jeong +3
The paper introduces a flow‑matching based generative framework for topology optimisation that incorporates physics‑guided intermediate states via a trajectory‑aware formulation, e…
Optimal Self-Distillation for Rectified Flow via Linear Probing
Saptarshi Roy, Debepsita Mukherjee, Pratik Patil
The paper investigates how to improve rectified flow generative models by optimally mixing teacher-generated velocity fields with true velocities, deriving a closed‑form mixing coe…
Synthesizing real-world distributions from high-dimensional Gaussian Noise with Fully Connected Neural Network
Joanna Komorniczak
The paper introduces a fully connected neural network that converts high‑dimensional Gaussian noise into synthetic tabular data resembling real datasets, using preprocessing, PCA,…
What Does Goodness Measure? A Likelihood-Ratio Account of Forward-Forward Learning
Paolo Giannitrapani
The paper explains that the Forward-Forward algorithm’s goodness measure is actually a likelihood‑ratio statistic under a generative model, and shows how different data distributio…
Discrete Diffusion Models: A Unified Framework from Tokenization to Generation
Ye Yuan, Weien Li, Rui Song +19
The paper proposes a unified framework for discrete denoising diffusion models that ties together tokenization, vocabulary design, and generation methods, showing how existing appr…
TreeSRNF: Square-Root Normal Fields for Generative Modelling of the Geometric and Structural Variability in Tree-like 3D Objects
Tahmina Khanam, Hamid Laga, Mohammed Bennamoun +4
The paper presents TreeSRNF, a mathematical framework that extends Square-Root Normal Fields to model both the geometry and branching structure of tree-like 3D objects, enabling an…
Heavy-Tailed Flow Matching via Random Clocks
Zhouhao Yang, Yezhen Wang, Kenji Kawaguchi +2
The paper introduces Heavy-Tailed Flow Matching via Random Clocks (HTFM), a method that models heavy‑tailed source distributions as mixtures of Gaussian flows conditioned on random…
ConFlow: Constraints-Guided Learning with Flow Matching for Motion Generation
Nutan Chen, Jianxiang Feng, Marvin Alles +1
ConFlow incorporates task constraints directly into the training of flow-matching models for robot motion generation, using differentiable barrier functions, a conditional Gaussian…
SinAE: A Single-Architecture Flow-Matching Autoencoder for Cross-Domain Atomic Systems
Yuxuan Ren, Fan Yang, Jianhua Yao +1
SinAE is a unified Transformer-based autoencoder that uses flow-matching decoding to reconstruct and generate atomic structures across molecules, crystals, and proteins with near-l…
The Geometry of Memorization: Finite-Time Spectral Sensitivity as a Diagnostic for Flow Matching Models
Shuchan Wang
The paper introduces Finite-Time Spectral Sensitivity (FTSS), a gradient‑free metric that tracks the RMS singular value of the state‑transition matrix in continuous‑time flow match…
AdaPCLA: Adaptive Prior-Calibrated Logit Adjustment for Long-Tailed Longitudinal EHR Generation
Shuai Cui, Chen Wenxuan, Wenjie Du +3
The paper introduces AdaPCLA, a framework that improves generative models for longitudinal electronic health records by adaptively adjusting logits to better represent rare disease…
ANGLE: Angular Neural Generative Learning via Engression
Rajdeep Pathak, Archi Roy, Tanujit Chakraborty
The paper introduces ANGLE, a lightweight deep generative model that learns the full conditional distribution of circular (angular) data given various covariates, enabling better p…
A Shortcut to Statistically Steady-State Turbulence with Flow Matching
Gianluca Galletti, Gerald Gutenbrunner, William Hornsby +5
The paper presents GyroFlow, a latent generative model that directly creates statistically steady‑state snapshots of gyrokinetic turbulence, avoiding the costly transient simulatio…
Generating synthetic evolution of turbulent flames with an experimental data-based spatiotemporal diffusion model
Amrit Tarur, Shivam Barwey
The paper presents a conditional diffusion model that generates synthetic spatiotemporal data of turbulent flames, reproducing experimental OH-PLIF and PIV measurements and enablin…
A Generative Model-Free Form Deformation Approach for the Generation of Mesh Motions with Applications to PDE
Gugliemo Padula, Artem Sinitsa, Gianluigi Rozza
The paper presents a topology‑agnostic framework that models 3D shape deformations as an ODE flow parameterized by a time‑dependent free‑form deformation lattice, and couples it wi…
One search, two signals: results blend meaning (embedding similarity, so papers that never use your words still surface) with keyword matches on titles, abstracts and summaries. Free, no sign-in needed.