Skip to main content
GameDev.net gamedev.net
96 sources covering this story

SCRIPT: Scalable Diffusion Policy with Multi-stage Training for Language-driven Physics-Based Humanoid Control

SCRIPT pushes language-driven humanoid control toward more stable, physically plausible motion. The system combines a diffusion policy with multi-stage training, then adds reinforcement learning to improve instruction following and motion quality in closed-loop simulation. For teams building embodied agents or character AI, the scaling results are the practical headline.

First reported 4 months, 1 week ago • ai humanoid control diffusion policy motion synthesis
Want to discuss what this means for developers?
Open discussion
PRIMARY SOURCE
arXiv cs.GR arXiv cs.GR

SCRIPT: Scalable Diffusion Policy with Multi-stage Training for Language-driven Physics-Based Humanoid Control

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

SC-MFJ: A Simple Haptic Quality Metric for Medical Image Segmentation

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

FontFusion: Enhancing Generative Text in Diffusion Models with Typographic Conditioning

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

The Invisible Hand of Physics: When Video Diffusion Models Know More Than They Show

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Balancing Image Compression and Generation with Bootstrapped Tokenization

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Monte Carlo Steklov Operators for Large-Scale Geometry Processing in the Wild

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Function2Scene: 3D Indoor Scene Layout from Functional Specifications

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Aggregating LLM-Based Weak Verifiers for Spatial Layout Generation

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

KV-Control: Parameter-Efficient K/V Injection for Trajectory-Controlled Text-to-Motion

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Oklch+: A Three-Parameter Extension of Oklab for Improved Color Difference Prediction

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

GS-NFS: Bandwidth-adaptive Streaming of Dynamic Gaussian Splats and Point Clouds

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

MeshFlow: Efficient Artistic Mesh Generation via MeshVAE and Flow-based Diffusion Transformer

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

PureLight: Learning Complex Luminaires with Light Tracing

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

SymTRELLIS: Symmetry-Enforced Voxel Latents for 3D Generation

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Homology-Preserving Dimensionality Reduction via Adaptive Mapper and Landmark Isomap

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Echo-Infinity: Learning Evolving Memory for Real-Time Infinite Video Generation

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Geometry Gaussians: Decoupling Appearance and Geometry in Gaussian Splatting

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Streami: An MPI Data-Parallel Library to Compute Field Lines on GPUs

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

PersistGS: Differentiable Physics for Object Permanence in 4D Gaussian Splatting

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Fewer, Better Frames: A Compute-Normalized Proof of Concept for Coherence-First World-Model Rendering with Model-Guided FSR4 Frame Generation

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

A Novel Procedural Generation for Level Design of Mansions and Dungeons

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

AvatarMix: Identity-Preserving Cross-Avatar Composition for Outfit Personalization

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

GS-ROR$^2$: Bidirectional-guided 3DGS and SDF for Reflective Object Relighting and Reconstruction

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

MPMWorlds: Material-Point-Method Simulations for Inferring and Extrapolating Physical Dynamics

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Composable function systems as a general-purpose rendering framework

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

KDH-CAD: Knowledge-data hybrid CAD learning under data scarcity

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Real-Time Physics Simulation with Dynamic Mesh-Gaussian Reconstructions

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

HiGS: A Hierarchical Rendering Architecture for Real-Time 3D Gaussian Splatting

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

PaintBench: Deterministic Evaluation of Precise Visual Editing

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Ultra Diffusion Poser: Diffusion-Based Human Motion Tracking From Sparse Inertial Sensors and Ranging-Based Between-Sensor Distances

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Advances in Neural 3D Mesh Texturing: A Survey

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

AlbedoEdit: Unified Instance-Level Video Editing with Albedo Guidance

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Subgrid Marching Tetrahedra

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Beyond Static Gaussians: An Empirical Investigation of Architectural Paradigms for Dynamic 3D Scene Reconstruction

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

SkelMo: Universal Skeletal Motion Generation for 3D Rigged Shapes

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Directed Distance Fields for Constant-Time Ray Queries on Gaussian Splatting

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Temporally-Aligned Evaluation for Audio-Driven Talking Head Generation

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Effective Multi-sensor Conditioning for Street-view Novel-view Synthesis

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Single-Line Drawing Generation via Semantics-Driven Optimization

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

ShapeLib: Designing a library of programmatic 3D shape abstractions with Large Language Models

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Optimizing 3D Gaussian Splatting via Point Cloud Upsampling

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

3DCodeBench: Benchmarking Agentic Procedural 3D Modeling Via Code

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

MidSurfNet: Learning Face Pairing for Mid-surface Abstraction of Thin-walled CAD Models

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

MultiAct: Text-to-Motion Generation from Composite Text via Tailored Attention Guidance

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

BijectiveRemesh: Maintaining Bijective Mappings for Data Transfer Across Remeshed Manifolds

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

LiftNav: Path Planning via Semantic Lifting in TSDF-Guided Gaussian Splatting

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Smaller and Faster 3DGS via Post-Training Dictionary Learning

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

DSD-GS: Dynamic-Static Decomposition of Gaussian Splatting for Efficient and High-Fidelity Dynamic Scene Reconstruction

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

SCALMU: Synthetically-trained Coupling of Adaptive Learned Multiplicative Updates for Hyperspectral-Multispectral Fusion

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

SWIM: Single-Instance Whole-Body Imitation for swiMming

4 months ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Cert-LAS: Toward Certified Model Ownership Verification for Text-to-Image Diffusion Models via Layer-Adaptive Smoothing

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

SuperVoxelGPT: Adaptive and Ordered 3D Tokenization for Autoregressive Shape Generation

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Before the Shutter: Aesthetic and Actionable Portrait Photography Planning in 3D Scenes

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

RAFI -- A Ray/Work Forwarding Infrastructure for Data Parallel Multi-Node/Multi-GPU Computing

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

FreeForm: Reduced-Order Deformable Simulation from Particle-Based Skinning Eigenmodes

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Auditing Training-Free 3D Shape Retrieval with Diffused Geodesic Moments

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

City-Mesh3R: Simulation-Ready City-Scale 3D Mesh Reconstruction from Multi-View Images

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Capsule: Efficient Player Isolation for Datacenters

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

NeuROK: Generative 4D Neural Object Kinematics

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Ambient-robust Inverse Rendering using Active RGB-NIR Imaging

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

The Principles of Diffusion Models

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

CLEAR-NeRF: Collinearity and Local-region Enhanced Accurate 3D Reconstruction in Unbounded Scenes

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Sketch2Motion: Text-driven 2D Sketch to 3D Animation via Diffusion-guided Skeleton Optimization

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Resolution-free neural surrogates for geometric parameterization and mapping with spatially varying fields

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Chirpy3D: Part-Aware Multi-View Diffusion for Creative Fine-Grained Object Generation

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

DinoComplete: 3D Shape Completion with Distilled Semantic Priors and State Space Models

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

PINNsur: Physics-Informed Neural Networks for PDEs on Curved Surfaces

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

BrickAnything: Geometry-Conditioned Buildable Brick Generation with Structure-Aware Tokenization

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Megakernel vs Wavefront GPU Path Tracing

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Garment Particles: A 2D--3D Symmetric Garment Representation for Generation and Editing

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Semantic-Aware Motion Encoding for Topology-Agnostic Character Animation

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

AnySurf: Any Surface Generation with Directed Edge

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Quantized Keys Steal Attention: Bias Correction for KV-Cache Compression in Video Diffusion

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Local MAP Sampling for Diffusion Models

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

{\Phi}-Noise: Training-Free Temporal Video Conditioning via Phase-Based Noise Manipulation

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

EMA: Effort Metric Attention for Anatomical Effort-Guided Human Motion Diffusion

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Curve Skeletonization in Continuous domain for Meshes and Point Clouds

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Progressively Projected Newton's Method

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Generating 3D models from sketches of human faces using a combined approach of Convolutional Neural Networks, Procedural Modeling, and Contour Mapping

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Squeezing Capacity from Multimodal Large Language Models for Subject-driven Generation

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Look Both Ways Before You Cross: Lifting Cross Fields From 2D Visual Priors

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Learning View-Dependent Splatting Kernels

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Snapshot Polarimetric Display Inverse Rendering

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

SRUG: Shadow-Guided Relightable Urban Scene with Generation Model

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

VectorArk: Learning Practical Image Vectorization with Rounded Polygon Representation

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Multi-view Consistent 3D Gaussian Head Avatars 'without' Multi-view Generation

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Depth Peeling for High-Fidelity Gaussian-Enhanced Surfel Rendering

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

MIND: Multi-Scale Intent Diffusion for Text-Driven Physics-Based Humanoid Control

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Appearance-Invariant Detection of Suggestive Motion via Laban Movement Descriptors

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

F-RNG: Feed-Forward Relightable Neural Gaussians

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Generalizable Vision-Language Few-Shot Adaptation with Predictive Prompts and Negative Learning

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

ChronoGS: Disentangling Invariants and Changes in Multi-Period Scenes

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Closing Trajectories: Equation-Free Cyclic Animation via Koopman Surrogates

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

Good Token Hunting: A Hitchhiker's Guide to Token Selection for Visual Geometry Transformers

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

DrawVideo: Generating Long Video from Storyboard Keyframe Sketches

4 months, 1 week ago Read source
arXiv cs.GR arXiv cs.GR 1% match

YASPS: A Symbolic Framework for Extensible, High-Performance IPC Simulation

4 months, 1 week ago Read source