Abstract
Reliable 3D spatial understanding is essential for autonomous navigation, obstacle avoidance, and scene reconstruction. While state-of-the-art learned depth estimation techniques achieve high accuracy in-distribution, they often generalize poorly to novel viewpoints and altitudes. This paper presents a geometrically derived, training-free depth estimation method using epipolar transfer with only two monocular images and camera pose estimates. By leveraging camera motion to synthesize a virtual stereo pair with a freely chosen baseline, our approach transforms temporal correspondence into a stereo triangulation task while mitigating geometric degeneracies inherent to direct two-view triangulation. Validated across outdoor drone flights (to a maximum range of approximately 90\,m) and indoor OptiTrack environments against LiDAR ground truth, the method achieves an indoor AbsRel of 0.092 and $δ< 1.25$ of 0.940, comparable to direct triangulation (AbsRel 0.073) while retaining valid depth over a larger fraction of challenging scenes, and substantially outperforms off-the-shelf learning-based baselines such as ZoeDepth (AbsRel 0.225) and Depth Anything V2 (AbsRel 0.570), which are not trained or fine-tuned for this domain, with no training data required.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Aggarwal, D., Dagadkhair, R., Srivastava, S., Denby, B., & Kochersberger, K. (2026). EpiTransfer: Sparse, Training-Free Long-Range Depth Estimation from Temporal Monocular Aerial Frames. https://omanscience.com/en/articles/epitransfer-sparse-training-free-long-range-depth-estimation-from-temporal-monocular-aerial-frames
MLA 9
Aggarwal, Diksha, et al. "EpiTransfer: Sparse, Training-Free Long-Range Depth Estimation from Temporal Monocular Aerial Frames." https://omanscience.com/en/articles/epitransfer-sparse-training-free-long-range-depth-estimation-from-temporal-monocular-aerial-frames.
Chicago (author–date)
Aggarwal, Diksha, Rutvik Dagadkhair, Sanjana Srivastava, Bradley Denby, and Kevin Kochersberger. 2026. "EpiTransfer: Sparse, Training-Free Long-Range Depth Estimation from Temporal Monocular Aerial Frames." https://omanscience.com/en/articles/epitransfer-sparse-training-free-long-range-depth-estimation-from-temporal-monocular-aerial-frames.
Harvard
Aggarwal, D., Dagadkhair, R., Srivastava, S., Denby, B. and Kochersberger, K. (2026) 'EpiTransfer: Sparse, Training-Free Long-Range Depth Estimation from Temporal Monocular Aerial Frames', Available at: https://omanscience.com/en/articles/epitransfer-sparse-training-free-long-range-depth-estimation-from-temporal-monocular-aerial-frames.
Vancouver
Aggarwal D, Dagadkhair R, Srivastava S, Denby B, Kochersberger K. EpiTransfer: Sparse, Training-Free Long-Range Depth Estimation from Temporal Monocular Aerial Frames. https://omanscience.com/en/articles/epitransfer-sparse-training-free-long-range-depth-estimation-from-temporal-monocular-aerial-frames
IEEE
D. Aggarwal, R. Dagadkhair, S. Srivastava, B. Denby, and K. Kochersberger, "EpiTransfer: Sparse, Training-Free Long-Range Depth Estimation from Temporal Monocular Aerial Frames," https://omanscience.com/en/articles/epitransfer-sparse-training-free-long-range-depth-estimation-from-temporal-monocular-aerial-frames.