Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Learning Global Motion with Compact Gaussians for Feed-Forward 4D Reconstruction

About

Dynamic scene reconstruction from monocular video remains a fundamental challenge in computer vision. Existing feed-forward methods predict 3D Gaussians pixel-wise for each frame, suffering from duplicated Gaussians and view-dependent biases that hinder effective learning of scene motion. We present C4G, a feed-forward 4D reconstruction framework built upon a compact set of timestamp-conditioned learnable Gaussian query tokens. Each token aggregates corresponding features across the full temporal context and decodes a 3D Gaussian whose position is modulated by the target timestamp, enabling globally coherent motion modeling without per-scene optimization. To capture fine-grained details, we further introduce a video diffusion model-based rendering enhancement module. Since our framework effectively aggregates features into Gaussians, we extend this capability to feature lifting, producing a 4D feature field that supports point tracking and dynamic scene understanding. C4G achieves strong novel-view synthesis performance using significantly fewer Gaussians and without requiring camera poses, while exhibiting stronger motion modeling and robustness to large temporal gaps.

Mungyeom Kim, Minkyeong Jeon, Honggyu An, Jaewoo Jung, Hyuna Ko, Jisang Han, Hyeonseo Yu, Donghwan Shin, Sunghwan Hong, Takuya Narihira, Kazumi Fukuda, Yuki Mitsufuji, Seungryong Kim• 2026

Related benchmarks

TaskDatasetResultRank
Novel View SynthesisNVIDIA
PSNR20.51
20
Video Object SegmentationDAVIS
IoU63.4
16
Novel View SynthesisTUM-D
PSNR19.52
10
Novel View SynthesisADT
PSNR22.35
10
Novel View SynthesisDyCheck--
6
Point TrackingDriveTrack 4
Success Rate (dt=2)55.51
4
Point TrackingADT 73
Accuracy (dt=2)93.98
4
Temporal-invariant Feature ExtractionDAVIS
< δ^0 Error2.2
4
Temporal-invariant Feature ExtractionRGB-Stacking
Displacement Error (< δ^0)4.9
4
Novel View SynthesisDAVIS
PSNR20.181
3
Showing 10 of 11 rows

Other info

Follow for update