Geo-Align applies RL with a perceptual reward derived from 3D camera trajectory estimation to improve controllability and fidelity in video generation without paired training data.
hub Canonical reference
MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion
Canonical reference. 88% of citing Pith papers cite this work as background.
34
Pith papers citing it
Background
88% of classified citations
abstract
Estimating geometry from dynamic scenes, where objects move and deform over time, remains a core challenge in computer vision. Current approaches often rely on multi-stage pipelines or global optimizations that decompose the problem into subtasks, like depth and flow, leading to complex systems prone to errors. In this paper, we present Motion DUSt3R (MonST3R), a novel geometry-first approach that directly estimates per-timestep geometry from dynamic scenes. Our key insight is that by simply estimating a pointmap for each timestep, we can effectively adapt DUST3R's representation, previously only used for static scenes, to dynamic scenes. However, this approach presents a significant challenge: the scarcity of suitable training data, namely dynamic, posed videos with depth labels. Despite this, we show that by posing the problem as a fine-tuning task, identifying several suitable datasets, and strategically training the model on this limited data, we can surprisingly enable the model to handle dynamics, even without an explicit motion representation. Based on this, we introduce new optimizations for several downstream video-specific tasks and demonstrate strong performance on video depth and camera pose estimation, outperforming prior work in terms of robustness and efficiency. Moreover, MonST3R shows promising results for primarily feed-forward 4D reconstruction.
hub tools
citation-role summary
background 7
method 1
citation-polarity summary
representative citing papers
NoPo4D is the first feed-forward system for dynamic 4D Gaussian splatting from unposed multi-view videos, using velocity decomposition supervised by optical flow and a bidirectional motion encoder.
Ground4D resolves temporal conflicts in feedforward 4D Gaussian reconstruction for off-road scenes via voxel-grounded temporal aggregation with intra-voxel softmax and surface normal regularization, outperforming prior methods on ORAD-3D and RELLIS-3D while generalizing zero-shot.
AirZoo is a new large-scale synthetic dataset for aerial 3D vision that improves state-of-the-art models on image retrieval, cross-view matching, and 3D reconstruction when used for fine-tuning.
Holo360D is the first large-scale dataset providing continuous panoramic sequences with accurately aligned high-completeness depth maps and meshes for training panoramic 3D reconstruction models.
Test-time constrained optimization incorporates priors into pre-trained multiview transformers via self-supervised losses and penalty terms to improve 3D reconstruction accuracy.
STAC compresses KV caches in streaming 3D reconstruction transformers via temporal token preservation with decayed attention, spatial voxel compression, and chunked multi-frame optimization, delivering 10x memory reduction and 4x faster inference at SOTA quality.
ZipMap achieves linear-time bidirectional 3D reconstruction by zipping image collections into a compact stateful representation via test-time training layers.
3AM integrates MUSt3R 3D features into SAM2 via a Feature Merger and FOV-aware sampling to deliver geometry-consistent video object segmentation from RGB alone, with large gains on wide-baseline datasets.
π³ is a feed-forward network with full permutation equivariance that outputs affine-invariant poses and scale-invariant local point maps without reference frames, reaching state-of-the-art on camera pose, depth, and dense reconstruction benchmarks.
UniT unifies online and offline 3D geometry perception via a Group Autoregressive Transformer that processes observation groups with anchor-free point map prediction and a scale-adaptive loss.
A training-free Spatio-Temporal Attention Chain framework accelerates 4D mesh generation 13x, improves quality, scales to 16x longer videos, and supports downstream tracking and camera estimation.
LongDPM introduces an overlap-aware chunk-based framework that registers and fuses local dynamic reconstructions to achieve coherent long-range 4D geometry and tracking from monocular video.
CoGE achieves state-of-the-art monocular geometric estimation in colonoscopy by training solely on simulated data via an illumination-aware Retinex-based module and a wavelet-based structure-aware module.
RetrieveVGGT enables constant-memory long-context streaming 3D reconstruction by retrieving relevant frames via query-key similarities in VGGT's first attention layer, outperforming StreamVGGT and others.
RigidFormer learns mesh-free rigid dynamics from point clouds using object-centric anchors, Anchor-Vertex Pooling, Anchor-based RoPE, and differentiable Kabsch alignment to enforce rigidity.
DeG models 3D Gaussians via learned octree density and uses VecSeq Sobol re-indexing to turn set generation into sequence modeling, claiming SOTA quality in single-image-to-3D.
Sat3R adapts Depth Anything V2 via RPC-aware metric depth fine-tuning to deliver satellite DSM reconstruction with 38% lower MAE than zero-shot baselines and over 300x speedup versus optimization methods.
The paper proposes ray-aware pointer memory with adaptive retain-or-replace updates to improve long-term stability and pose accuracy in streaming 3D reconstruction.
Finetuning 3D foundation models on simulated sparse subsets from MegaDepth-X produces robust reconstructions from extremely sparse, noisy internet photos while preserving performance on dense benchmarks.
Vista4D re-synthesizes dynamic videos from new viewpoints by grounding them in a 4D point cloud built with static segmentation and multiview training.
The paper proposes a problem-driven taxonomy for feed-forward 3D scene modeling that groups methods by five core challenges: feature enhancement, geometry awareness, model efficiency, augmentation strategies, and temporal-aware modeling.
SelfEvo enables pretrained 4D perception models to self-improve on unlabeled videos via self-distillation, delivering up to 36.5% relative gains in video depth estimation and 20.1% in camera estimation across eight benchmarks.
SceneScribe-1M is a new dataset of 1 million videos with semantic text, camera parameters, dense depth, and consistent 3D point tracks to support monocular depth estimation, scene reconstruction, point tracking, and text-to-video synthesis.
citing papers explorer
-
Geo-Align: Video Generation Alignment via Metric Geometry Reward
Geo-Align applies RL with a perceptual reward derived from 3D camera trajectory estimation to improve controllability and fidelity in video generation without paired training data.
-
No Pose, No Problem in 4D: Feed-Forward Dynamic Gaussians from Unposed Multi-View Videos
NoPo4D is the first feed-forward system for dynamic 4D Gaussian splatting from unposed multi-view videos, using velocity decomposition supervised by optical flow and a bidirectional motion encoder.
-
Ground4D: Spatially-Grounded Feedforward 4D Reconstruction for Unstructured Off-Road Scenes
Ground4D resolves temporal conflicts in feedforward 4D Gaussian reconstruction for off-road scenes via voxel-grounded temporal aggregation with intra-voxel softmax and surface normal regularization, outperforming prior methods on ORAD-3D and RELLIS-3D while generalizing zero-shot.
-
AirZoo: A Unified Large-Scale Dataset for Grounding Aerial Geometric 3D Vision
AirZoo is a new large-scale synthetic dataset for aerial 3D vision that improves state-of-the-art models on image retrieval, cross-view matching, and 3D reconstruction when used for fine-tuning.
-
Holo360D: A Large-Scale Real-World Dataset with Continuous Trajectories for Advancing Panoramic 3D Reconstruction and Beyond
Holo360D is the first large-scale dataset providing continuous panoramic sequences with accurately aligned high-completeness depth maps and meshes for training panoramic 3D reconstruction models.
-
Learning 3D Reconstruction with Priors in Test Time
Test-time constrained optimization incorporates priors into pre-trained multiview transformers via self-supervised losses and penalty terms to improve 3D reconstruction accuracy.
-
STAC: Plug-and-Play Spatio-Temporal Aware Cache Compression for Streaming 3D Reconstruction
STAC compresses KV caches in streaming 3D reconstruction transformers via temporal token preservation with decayed attention, spatial voxel compression, and chunked multi-frame optimization, delivering 10x memory reduction and 4x faster inference at SOTA quality.
-
ZipMap: Linear-Time Stateful 3D Reconstruction via Test-Time Training
ZipMap achieves linear-time bidirectional 3D reconstruction by zipping image collections into a compact stateful representation via test-time training layers.
-
3AM: 3egment Anything with Geometric Consistency in Videos
3AM integrates MUSt3R 3D features into SAM2 via a Feature Merger and FOV-aware sampling to deliver geometry-consistent video object segmentation from RGB alone, with large gains on wide-baseline datasets.
-
$\pi^3$: Permutation-Equivariant Visual Geometry Learning
π³ is a feed-forward network with full permutation equivariance that outputs affine-invariant poses and scale-invariant local point maps without reference frames, reaching state-of-the-art on camera pose, depth, and dense reconstruction benchmarks.
-
UniT: Unified Geometry Learning with Group Autoregressive Transformer
UniT unifies online and offline 3D geometry perception via a Group Autoregressive Transformer that processes observation groups with anchor-free point map prediction and a scale-adaptive loss.
-
Fast 4D Mesh Generation by Spatio-Temporal Attention Chains
A training-free Spatio-Temporal Attention Chain framework accelerates 4D mesh generation 13x, improves quality, scales to 16x longer videos, and supports downstream tracking and camera estimation.
-
LongDPM: Overlap-Aware 4D Reconstruction from Long Monocular Videos
LongDPM introduces an overlap-aware chunk-based framework that registers and fuses local dynamic reconstructions to achieve coherent long-range 4D geometry and tracking from monocular video.
-
CoGE: Sim-to-Real Online Geometric Estimation for Monocular Colonoscopy
CoGE achieves state-of-the-art monocular geometric estimation in colonoscopy by training solely on simulated data via an illumination-aware Retinex-based module and a wavelet-based structure-aware module.
-
Attention Itself Could Retrieve.RetrieveVGGT: Training-Free Long Context Streaming 3D Reconstruction via Query-Key Similarity Retrieval
RetrieveVGGT enables constant-memory long-context streaming 3D reconstruction by retrieving relevant frames via query-key similarities in VGGT's first attention layer, outperforming StreamVGGT and others.
-
RigidFormer: Learning Rigid Dynamics using Transformers
RigidFormer learns mesh-free rigid dynamics from point clouds using object-centric anchors, Anchor-Vertex Pooling, Anchor-based RoPE, and differentiable Kabsch alignment to enforce rigidity.
-
Generative 3D Gaussians with Learned Density Control
DeG models 3D Gaussians via learned octree density and uses VecSeq Sobol re-indexing to turn set generation into sequence modeling, claiming SOTA quality in single-image-to-3D.
-
Sat3R: Satellite DSM Reconstruction via RPC-Aware Depth Fine-tuning
Sat3R adapts Depth Anything V2 via RPC-aware metric depth fine-tuning to deliver satellite DSM reconstruction with 38% lower MAE than zero-shot baselines and over 300x speedup versus optimization methods.
-
Ray-Aware Pointer Memory with Adaptive Updates for Streaming 3D Reconstruction
The paper proposes ray-aware pointer memory with adaptive retain-or-replace updates to improve long-term stability and pose accuracy in streaming 3D reconstruction.
-
Long-tail Internet photo reconstruction
Finetuning 3D foundation models on simulated sparse subsets from MegaDepth-X produces robust reconstructions from extremely sparse, noisy internet photos while preserving performance on dense benchmarks.
-
Vista4D: Video Reshooting with 4D Point Clouds
Vista4D re-synthesizes dynamic videos from new viewpoints by grounding them in a 4D point cloud built with static segmentation and multiview training.
-
Feed-Forward 3D Scene Modeling: A Problem-Driven Perspective
The paper proposes a problem-driven taxonomy for feed-forward 3D scene modeling that groups methods by five core challenges: feature enhancement, geometry awareness, model efficiency, augmentation strategies, and temporal-aware modeling.
-
Self-Improving 4D Perception via Self-Distillation
SelfEvo enables pretrained 4D perception models to self-improve on unlabeled videos via self-distillation, delivering up to 36.5% relative gains in video depth estimation and 20.1% in camera estimation across eight benchmarks.
-
SceneScribe-1M: A Large-Scale Video Dataset with Comprehensive Geometric and Semantic Annotations
SceneScribe-1M is a new dataset of 1 million videos with semantic text, camera parameters, dense depth, and consistent 3D point tracks to support monocular depth estimation, scene reconstruction, point tracking, and text-to-video synthesis.
-
OVGGT: O(1) Constant-Cost Streaming Visual Geometry Transformer
OVGGT achieves constant O(1) memory and compute for streaming 3D geometry reconstruction by using FFN-residual-based KV cache compression and dynamic anchor protection, matching state-of-the-art accuracy on long sequences.
-
Densemarks: Learning Canonical Embeddings for Human Heads Images via Point Tracks
DenseMarks learns a canonical 3D embedding space for human head images by training a Vision Transformer with contrastive loss on pairwise point tracks from in-the-wild videos, plus landmark and segmentation supervision.
-
Diff4Splat: Controllable 4D Scene Generation with Latent Dynamic Reconstruction Models
A feed-forward video latent transformer that predicts time-varying 3D Gaussian primitives from one image to produce controllable 4D scenes with appearance, geometry, and motion.
-
Streaming 4D Visual Geometry Transformer
A causal transformer with key-value caching and distillation from a bidirectional VGGT model enables efficient online 4D geometry reconstruction from videos.
-
VLM-3R: Vision-Language Models Augmented with Instruction-Aligned 3D Reconstruction
VLM-3R augments VLMs with implicit 3D tokens from monocular video via geometry encoding and 200K+ 3D reconstructive QA pairs, plus a new 138K-pair temporal benchmark, to support spatial and embodied reasoning.
-
IVGT: Implicit Visual Geometry Transformer for Neural Scene Representation
IVGT implicitly models continuous neural scene representations from pose-free multi-view images to enable coherent surface extraction, novel view synthesis, and related 3D tasks via SDF and color prediction.
-
WildPose: A Unified Framework for Robust Pose Estimation in the Wild
WildPose unifies feedforward 3D features from MASt3R with differentiable bundle adjustment for robust monocular pose estimation across dynamic, static, and low-ego-motion scenes.
-
ViPE: Video Pose Engine for 3D Geometric Perception
ViPE estimates camera intrinsics, motion, and dense near-metric depth from uncalibrated videos, outperforming baselines on TUM and KITTI while releasing annotations for 96M frames across real and generated videos.
-
LychSim: A Controllable and Interactive Simulation Framework for Vision Research
LychSim introduces a controllable simulation platform on Unreal Engine 5 with Python API, procedural generation, and LLM integration for vision research tasks.
-
DINO_4D: Semantic-Aware 4D Reconstruction
DINO_4D uses frozen DINOv3 features to inject semantic awareness into 4D dynamic scene reconstruction, improving tracking accuracy and completeness on benchmarks while preserving O(T) complexity.