Schedule

We are excited to host invited talks from a series of fantastic speakers who are experts in differentiable vision, graphics, physics, and allied areas. Additionally, we have five contributed talks chosen from the submissions received.

The workshop will be held on 11 December 2020 (Friday). All times below are listed in Eastern Standard Time (Toronto/New York)

All papers will be presented as posters in a virtual session organized via gather.town. Check here for a guide to our posters.

0945 - 1000

Organizers

1000 - 1025

1025 - 1030

Live Q&A

1030 - 1055

1055 - 1100

Live Q&A

1102 - 1131

1132 - 1137

Live Q&A

1138 - 1153

1154 - 1159

Live Q&A

1200 - 1213

Oral 1: phiflow - A differentiable PDE solving framework for deep learning via physical simulations

1213 - 1223

Oral 2: Differentiable HDR image synthesis using multi-exposure images

1223 - 1235

Oral 3: DELUCA - Differentiable control library - environments, methods, and benchmarking

1235 - 1244

Oral 4: Blendshape-augmented facial action units detection

1244 - 1257

Oral 5: Inverse articulated-body dynamics from video via sequential Monte-Carlo

1258 - 1308

Live Q&A for Oral 1 through 5

1310 - 1335

1336 - 1341

Live Q&A

1342 - 1413

1414 - 1419

Live Q&A

1420 - 1439

1440 - 1445

Live Q&A

1500 - 1525

1525 - 1530

Live Q&A

1530 - 1645

Panel discussion: The promises and perils of differentiable methods YouTube link

1645 - 1800

Poster session (on gather.town)