DriveDreamer: Towards Real-world-driven World Models for Autonomous Driving

cs.RO 2026-07 conditional novelty 7.0 of 10

DreamerV3's imagined rollouts are insensitive to friction changes that cause real gait collapse, revealing that world models extrapolate kinematically rather than dynamically.

cs.CV 2025-08 conditional novelty 7.0 of 10

LSD-3D generates explicit, 3D-consistent driving scenes by combining a generated proxy mesh with geometry-grounded distillation from a 2D diffusion model.

cs.CV 2024-12 conditional novelty 7.0 of 10

A joint diffusion framework trains a Stable Diffusion generator and a semantic occupancy perception model together, so each task improves the other, producing text-conditional RGB-occupancy pairs.

cs.CV 2026-08 conditional novelty 6.0 of 10

Khora decouples world-state evolution from visual rendering through a shared STBoard and fixed-dimensional per-view renderers, enabling inference-time addition and removal of agents without retraining.

cs.CV 2026-08 conditional novelty 6.0 of 10

GSRAIN fuses measured-raindrop-calibrated Gaussian streaks/haze with a diffusion-based rainy-appearance transfer in 3D Gaussian Splatting scenes, enabling 0–13 mm/h rainfall control for autonomous-driving testing.

cs.RO 2026-07 conditional novelty 6.0 of 10

A driving planner that predicts a future-ego-trajectory latent and retrieves executable trajectories from a fixed memory reaches 91.3 PDMS on NAVSIM v1 without reconstructing the scene.

cs.CV 2026-07 conditional novelty 6.0 of 10

Adding persistently updated, supervised world-state register tokens to streaming multi-agent diffusion improves cross-agent consistency and visual quality in two-agent Minecraft generation.

cs.GR 2026-07 conditional novelty 6.0 of 10

A feed-forward model reconstructs a layered, simulation-ready 3D Gaussian world from multi-view driving video in ~1.5 s, with quality approaching per-scene optimized reconstruction.

cs.CV 2025-12 conditional novelty 6.0 of 10

A vision-language model writes a simulation program—grounded by segmentation and 3D tools—that predicts physically plausible futures from an image and caption, outperforming video generators on modified benchmarks.

cs.CV 2025-11 conditional novelty 6.0 of 10

FSU-QA is a new nuScenes-based VQA benchmark for foresight reasoning; current VLMs score about 41-49% accuracy, and fine-tuning an 8B model on it outperforms GPT-5 and Gemini on the benchmark.

cs.CV 2025-06 conditional novelty 6.0 of 10

An autoregressive diffusion world model jointly generates the next camera frame and a multi-step trajectory, enabling long videos and real-time planning for autonomous driving.

cs.CV 2025-06 conditional novelty 6.0 of 10

A pipeline that restores corrupted novel-view videos with a video diffusion model and jointly denoises multiple viewpoints to improve 3D scene exploration from a single image.

cs.RO 2025-06 conditional novelty 6.0 of 10

A hierarchical leader-follower Gaussian world model improves multi-task bimanual manipulation success rates over prior single-arm-based methods.

cs.CV 2025-06 reject novelty 6.0 of 10

TARDIS is a transformer world model trained on STRIDE, a graph-structured street-view dataset, with claimed abilities in controllable image generation, georeferencing, self-driving actions, and temporal simulation.

cs.CV 2025-06 conditional novelty 6.0 of 10

A three-stage hybrid pipeline uses an intermediate layered world representation to refine simulator-rendered driving scenes into realistic, controllable images and videos.

cs.CV 2025-06 conditional novelty 6.0 of 10

MeWM combines a GPT-style policy, a diffusion tumor dynamics model, and a survival analysis heuristic to simulate post-treatment tumor appearance and select TACE treatment plans, improving physician F1-score by 13 points.

cs.CV 2025-05 conditional novelty 6.0 of 10

ProphetDWM is a one-stage diffusion world model that jointly predicts future driving video and low-level actions from a current frame and a short action sequence.

cs.CV 2025-05 conditional novelty 6.0 of 10

A pose-control module using self-supervised depth and forward plus inverse warping losses improves camera alignment in diffusion and autoregressive world models.

cs.CV 2025-04 conditional novelty 6.0 of 10

NoiseController decomposes initial diffusion noise into scene-level foreground/background and shared/residual components, then collaborates them across views and frames, improving multi-view video consistency on nuScenes.

cs.CV 2025-02 conditional novelty 6.0 of 10

SCMA trains a policy-agnostic observation denoiser, using a pre-trained world model as a clean-distribution reference, and shows improved visual RL performance under distractions.

cs.RO 2025-01 conditional novelty 6.0 of 10

AdaWM selectively finetunes either the dynamics model or the policy of a pretrained world-model driving agent according to which mismatch dominates, improving success rates in CARLA.

cs.CV 2024-12 conditional novelty 6.0 of 10

DreamDrive generates 3D-consistent driving videos from a single image by lifting diffusion-generated reference frames into a hybrid static and dynamic 4D Gaussian scene.

cs.CV 2024-12 conditional novelty 6.0 of 10

A 1B-parameter autoregressive world model generates over 40 seconds of controllable driving video with near-SOTA FVD, but key claims rest on inconsistent and cross-paper comparisons.

cs.CV 2024-12 conditional novelty 6.0 of 10

DriveEditor uses depth-aware 3D bounding box projection and single-reference appearance cues to reposition, replace, remove, and insert objects in driving videos with a single diffusion framework.

cs.CV 2024-12 conditional novelty 6.0 of 10

VLM-AD uses GPT-4o-generated reasoning and action annotations as auxiliary supervision to improve end-to-end autonomous driving planning without VLM inference.

cs.CV 2024-12 conditional novelty 6.0 of 10

An end-to-end, non-autoregressive 3D occupancy world model warps dynamic voxels via predicted flow, moves static voxels by pose, and uses image-based rendering supervision, achieving state-of-the-art results on three ...

cs.CV 2024-12 conditional novelty 6.0 of 10

StreetCrafter conditions a video diffusion model on LiDAR point cloud renderings to synthesize controllable street views, and distills it into a real-time 3D Gaussian representation.

cs.CV 2024-12 conditional novelty 6.0 of 10

GEM generates controllable future RGB and depth ego-vision frames, conditioned on ego-trajectories, sparse object tokens, and human poses, across driving, egocentric, and drone domains.

cs.CV 2024-12 conditional novelty 6.0 of 10

Doe-1 unifies perception, prediction, and planning in autonomous driving into a single autoregressive next-token generation model over image, text, and action tokens.

cs.CV 2024-12 conditional novelty 6.0 of 10

Alternating semantic-map tokens and RGB tokens during autoregressive video continuation improves long-term consistency and FVD for driving videos.

cs.CV 2024-12 conditional novelty 6.0 of 10

A generation-reconstruction pipeline with a diffusion enhancer trained on simulated degradations enables off-trajectory camera rendering in driving scenes.

cs.CV 2024-12 conditional novelty 6.0 of 10

A framework that jointly generates multi-view camera images and LiDAR point clouds for driving scenes, with cross-modal consistency, improving on prior single-modality generators.

cs.CV 2024-11 conditional novelty 6.0 of 10

MagicDrive-V2 shows that matching geometric-control encoders to a 3D VAE's 4x temporal compression enables controllable, high-resolution, long multi-view driving video generation.

cs.CV 2024-11 conditional novelty 6.0 of 10

GenEx generates consistent 360-degree exploration videos from a single egocentric panorama and shows that LLM agents using these imagined views make more accurate decisions.

cs.RO 2024-11 conditional novelty 6.0 of 10

DrivingSphere combines occupancy-based 4D world generation with video diffusion to create a closed-loop simulation environment for autonomous driving evaluation.

cs.CV 2024-11 conditional novelty 6.0 of 10

EgoVid-5M is a 5M-clip curated egocentric video dataset with action and kinematic annotations, and EgoDreamer generates egocentric videos from text and camera control signals.

cs.CV 2026-08 conditional novelty 5.0 of 10

An agent-centric reframing of world modeling, replacing physical state prediction with 'information transitions' organized into six proxy functions and three empowerment levels.

cs.CV 2026-01 conditional novelty 5.0 of 10

A unified VLM for autonomous driving that couples trajectory planning with future-frame image generation improves open- and closed-loop planning metrics on Bench2Drive and nuScenes.

cs.NI 2025-07 conditional novelty 5.0 of 10

The paper proposes MobiWorld, a diffusion-based controllable world model for mobile networks, and reports a preliminary energy-saving optimization case study.

cs.RO 2025-07 conditional novelty 5.0 of 10

A Real2Sim2Real framework that aligns simulator dynamics via differentiable parameter fitting and renders photorealistic policy-training videos with a diffusion model, improving real-world manipulation success.

cs.CV 2025-06 reject novelty 5.0 of 10

A hierarchical coarse-to-fine diffusion transformer with cross-granularity distillation improves long-term driving video prediction, but the reported gains may be inflated by future-derived text prompts and a selected...

cs.RO 2025-01 conditional novelty 5.0 of 10

PIWM, an individual-vehicle world model with self-attention interaction modeling and trajectory-prediction representation learning, beats DreamerV3 and model-free RL on INTERACTION-based driving benchmarks.

cs.CV 2024-12 conditional novelty 5.0 of 10

A world model operating on 3D Gaussians forecasts the current occupancy from the previous frame and current RGB, improving mIoU by about 2 points on nuScenes without meaningful added latency.

cs.CV 2024-12 conditional novelty 5.0 of 10

DrivePhysica adds coordinate alignment, 3D instance flow, and box-coordinate guidance to a diffusion world model, achieving state-of-the-art FID/FVD on nuScenes and improving StreamPETR NDS by 3.6 points when mixed wi...

cs.RO 2024-12 conditional novelty 5.0 of 10

A reactive closed-loop driving simulator that uses a diffusion renderer with retrieval from real recordings, plus a nuPlan behavioral controller, to generate sensor images in response to an end-to-end driving model's actions.

cs.CV 2024-11 conditional novelty 5.0 of 10

ReconDreamer fine-tunes a driving world model as an online restorer and progressively expands novel-trajectory training data, reporting first-time effective rendering of multi-lane shifts in driving scenes.

cs.CV 2024-12 conditional novelty 4.0 of 10

A diffusion transformer with joint attention over all camera views and frames, a lightweight BEV controller, and a re-weighted object loss reports state-of-the-art multi-view driving video generation (FVD 37.8 on nuScenes).

cs.CV 2024-11 reject novelty 4.0 of 10

VAMP scores generated videos by combining per-object color, shape, and texture consistency with centroid velocity and acceleration smoothness, but its physics-based and human-aligned claims are not supported by the re...