Program

(Click on the presentation title to see the slides).

March 9th

13:00 – Welcome

13:25 – Opening Words – Emmanuel Baccelli, on behalf of ECDF & Inria Berlin

13:30 – Session 1

• Synthesizing 3D Consistent Human Heads via 3DGS GANs
  Florian Barthel, Humboldt-Universität zu Berlin & Fraunhofer HHI.
• Capturing Polysemanticity with PRISM
  Laura Kopf, Technische Universität Berlin.
• Fusion-based RAM Optimizations for TinyML
  Zhaolan Huang, Freie Universität Berlin & Inria Berlin.

15:00 – Coffee Break

15:30 – Session 2

• Learning Depth with Event Cameras (DERD-Net)
  Guillermo Gallego, TU Berlin & ECDF.
• Correcting Misinterpretations of Additive Models
  Benedict Clark, National Metrology Institute (PTB).
• Rethinking Explanation Evaluation under the Retraining Scheme
  Yi Cai and Gerhard Wunder, Freie Universität Berlin.

17:00 – Wrap-up.

March 10th

13:00 – Welcome

13:30 – Session 3

• Efficient Quadratic Corrections for Frank-Wolfe Algorithms
  Sebastian Pokutta, Zuse-Institut Berlin.
• Collapsing Taylor Mode Automatic Differentiation
  Tim Siebert, Humboldt-Universität zu Berlin.
• Internal MLLM Representations for Efficient Fine-grained Visual Question Answering
  Hanno Gottschalk, Technische Universität Berlin.

15:00 – Coffee Break

15:30 – Session 4

• tinyML Multiobjective Optimization for Urban Infrastructure Applications
  Vipin Singh and Felix Bießmann, BHT & ECDF.
• IndEgo: A Dataset of Industrial Scenarios and Collaborative Work for Egocentric Assistants
  Vivek Chavan, Fraunhofer IPK & TU Berlin.
• Low-Rank Training and Low-Rank Adapters
  Timon Klein, Otto-von-Guericke-Universität Magdeburg.

17:00 – Closing words.