Workshop
2nd Monocular Depth Estimation Challenge
Jaime Spencer · C. Stella Qian · Chris Russell · Simon Hadfield · Erich Graf · Wendy Adams · Andrew J. Schofield · James Elder · Richard Bowden
West 208 - 209
Sun 18 Jun, 8:30 a.m. PDT
Keywords: Single-view 3D
Monocular depth estimation (MDE) is an important low-level vision task, with application in fields such as augmented reality, robotics and autonomous vehicles. Recently, there has been an increased interest in self-supervised systems capable of predicting the 3D scene structure without requiring ground-truth LiDAR training data. Automotive data has accelerated the development of these systems, thanks to the vast quantities of data, the ubiquity of stereo camera rigs and the mostly-static world. However, the evaluation process has also remained focused on only the automotive domain and has been largely unchanged since its inception, relying on simple metrics and sparse LiDAR data.
This workshop seeks to answer the following questions:
1. How well do networks generalize beyond their training distribution relative to humans?
2. What metrics provide the most insight into the model’s performance? What is the relative weight of simple cues, e.g. height in the image, in networks and humans?
3. How do the predictions made by the models differ from how humans perceive depth? Are the failure modes the same?
The workshop will therefore consist of two parts: invited keynote talks discussing current developments in MDE and a challenge organized around a novel benchmarking procedure using the SYNS dataset.
Schedule
Sun 8:30 a.m. - 8:35 a.m.
|
Introduction
(
Talk
)
>
|
🔗 |
Sun 8:35 a.m. - 9:15 a.m.
|
Oisin Mac Aodha – Advancing Monocular Depth Estimation
(
Keynote
)
>
|
🔗 |
Sun 9:15 a.m. - 9:40 a.m.
|
Jaime Spencer – The Monocular Depth Estimation Challenge
(
Keynote
)
>
|
🔗 |
Sun 9:40 a.m. - 9:55 a.m.
|
Linh Trinh – Challenge Winner (Self-Supervised)
(
Challenge Participant
)
>
|
🔗 |
Sun 9:55 a.m. - 10:10 a.m.
|
Wei Yin – Challenge Winner (Supervised)
(
Challenge Participant
)
>
|
🔗 |
Sun 10:10 a.m. - 10:40 a.m.
|
Break
|
🔗 |
Sun 10:40 a.m. - 11:20 a.m.
|
Daniel Cremers – From Monocular Depth Estimation to 3D Scene Reconstruction
(
Keynote
)
>
|
🔗 |
Sun 11:20 a.m. - 12:00 p.m.
|
Alex Kendall – Building the Foundation Model for Embodied AI
(
Keynote
)
>
|
🔗 |