The Monocular Depth Estimation Challenge

Spencer, Jaime; Qian, Chaoxu; Russell, Chris; Hadfield, Simon; Graf, Erich W.; Adams, Wendy J.; Schofield, Andrew J.; Elder, James H.; Bowden, Richard; Cong, Heng; Mattoccia, Stefano; Poggi, Matteo; Suri, Zeeshan Khan; Tang, Yang; Tosi, Fabio; Wang, Hao; Zhang, Youmin; Zhang, Yusheng; Zhao, Chaoqiang

doi:10.1109/wacvw58289.2023.00069

Cited by 15 publications

(7 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This represents a relative improvement of 27.62% in F-Score (13.72% -Baseline) and 18% in AbsRel (29.66% -OPDAI) w.r.t. the first edition of the challenge [77]. The top-performing self-supervised method was Team imec-IDLab-UAntwerp, which leveraged improved pretrained encoders and deformable decoder convolutions.…”

Section: Quantitative Resultsmentioning

confidence: 99%

“…The Monocular Depth Estimation Challenge series [77]-the focus of this paper-is based on the MonoDepth Benchmark [78], which provided fair evaluations and implementations of recent SotA self-supervised MDE algorithms. Our focus lies on zero-shot generalization to a wide diversity of scenes.…”

Section: Related Workmentioning

confidence: 99%

“…Metrics. We follow the metrics used in the first edition of the challenge [77], categorized as image-/pointcloud-/edgebased. Image-based metrics represent the most common metrics (MAE, RMSE, AbsRel) computed using pixel-wise comparisons between the predicted and ground-truth depth map.…”

Section: The Monocular Depth Estimation Challengementioning

confidence: 99%

“…The first edition of MDEC [77] focused on benchmarking self-supervised approaches. The submissions outperformed the baseline [25,78] in all image-based metrics (Ab-sRel, MAE, RMSE), but provided slightly inferior pointcloud reconstructions [62] (F-Score).…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

2023 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW)

2023

View full text Add to dashboard Cite

This paper discusses the results for the second edition of the Monocular Depth Estimation Challenge (MDEC). This edition was open to methods using any form of supervision, including fully-supervised, self-supervised, multi-task or proxy depth. The challenge was based around the SYNS-Patches dataset, which features a wide diversity of environments with high-quality dense ground-truth. This includes complex natural environments, e.g. forests or fields, which are greatly underrepresented in current benchmarks.The challenge received eight unique submissions that outperformed the provided SotA baseline on any of the pointcloud-or image-based metrics. The top supervised submission improved relative F-Score by 27.62%, while the top self-supervised improved it by 16.61%. Supervised submissions generally leveraged large collections of datasets to improve data diversity. Self-supervised submissions instead updated the network architecture and pretrained backbones. These results represent a significant progress in the field, while highlighting avenues for future research, such as reducing interpolation artifacts at depth boundaries, improving self-supervised indoor performance and overall natural image accuracy.

show abstract

Section: Quantitative Resultsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: The Monocular Depth Estimation Challengementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

2023 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW)

2023

View full text Add to dashboard Cite

show abstract

“…The estimation of depth from a single image is an illposed problem, and numerous methods have been proposed to address it [26,27,28,29]. Initially, supervised learning dominated the utilized training approaches, with Egin et al [2] proposing comprehensive metrics.…”

Section: Self-supervised Monocular Depth Estimationmentioning

confidence: 99%

MonoViT-3D: Self-Supervised Monocular Depth Estimation from 3D Geometric Constraints using Point Cloud Projection

Wen,

Tang,

Sun

et al. 2024

Preprint

View full text Add to dashboard Cite

Self-supervised monocular depth estimation leverages geometric constraints to create pseudolabels, eliminating the need for ground-truth data during training. In the field of 3D geometric constraints, the optical flow and iterative closest point methods are widely used to extract geometric constraints. However, the application of the existing 3D geometric constraint methods in self-supervised learning tasks increases network complexity by introducing additional submodules and iterative loops. In this paper, we propose a point cloud projection method that replaces the iterative closest point algorithm. Following depth prediction, our method utilizes the difference of 3D gaussian measure to extract 3D point cloud structures and the difference of 2D gaussian measure for the 2D structures derived from synthesized multiview surfaces. By projecting a 3D structure onto a 2D image, the errors between the target point cloud image and the synthesized views are estimated to enhance the training process of self-supervised learning. 3D geometric constraints form a point cloud optimization problem in our approach, requiring deep prediction networks with strong receptive fields for global features. Benefiting from the global receptive field of transformers, we use a vision transformer for self-supervised learning. Furthermore, we propose an encoder block to more effectively learn the patterns of 3D geometric constraints. Experiments show that we achieve strong performance on the KITTI dataset, achieving 0.095, 0.691, 4.243, and 0.171 Abs Rel, Sq Rel, RMSE, and RMSE log values for monocular depth estimation, respectively. Moreover, the model outperforms the comparison models in 3D generation tasks conducted on the Make3D dataset. All code is available at https://github.com/weentiaan/MonoViT-3D.

show abstract