DELTA: Dense Efficient Long-range
3D Tracking for Any video

Tuan Duc Ngo^1,2 Peiye Zhuang¹ Chuang Gan² Evangelos Kalogerakis^2,3
Sergey Tulyakov¹ Hsin-Ying Lee¹ Chaoyang Wang^1
1 Snap Inc ² UMass Amherst ³ TU Crete

[Paper]      [Arxiv]      [Code (soon)]     [BibTeX]

DELTA captures dense, long-range, 3D trajectories from casual videos in a feed-forward manner.

Abstract

Tracking dense 3D motion from monocular videos remains challenging, particularly when aiming for pixel-level precision over long sequences. We introduce DELTA, a novel method that efficiently tracks every pixel in 3D space, enabling accurate motion estimation across entire videos. Our approach leverages a joint global-local attention mechanism for reduced-resolution tracking, followed by a transformer-based upsampler to achieve high-resolution predictions. Unlike existing methods, which are limited by computational inefficiency or sparse tracking, DELTA delivers dense 3D tracking at scale, running over 8x faster than previous methods while achieving state-of-the-art accuracy. Furthermore, we explore the impact of depth representation on tracking performance and identify log-depth as the optimal choice. Extensive experiments demonstrate the superiority of DELTA on multiple benchmarks, achieving new state-of-the-art results in both 2D and 3D dense tracking tasks. Our method provides a robust solution for applications requiring fine-grained, long-term motion tracking in 3D space.

Motivation

Existing motion prediction methods struggle with short-term, sparse predictions and often fail to deliver accurate 3D motion estimations while optimization-based approaches require substantial time to process a single video. We are the first method capable of efficiently tracking every pixel in 3D space over hundreds of frames from monocular videos, and achieves state-of-the-art accuracy on 3D tracking benchmarks.

Method

DELTA takes RGB-D videos as input and achieves efficient dense 3D tracking using a coarse-to-fine strategy, beginning with coarse tracking through a spatio-temporal attention mechanism at reduced resolution, followed by an attention-based upsampler for high-resolution predictions.

More qualitative results

Comparison with previous work

More results of 3D dense tracking can be found here.

Application: Non-rigid structure from motion

We first densely track pixels across multiple keyframes in the video to obtain pairwise correspondences. Using these correspondences, we jointly estimate per-keyframe depth maps and camera poses through the Global Alignment in DUSt3R and MonST3R.

More results of video pose estimation can be found here.

Application: Consistent video editting in 3D space