51³Ô¹ÏÍø

The DoC is committed to supporting our researchers and academics in their endeavors to explore uncharted territories and solve complex challenges.

The Department of Computing at 51³Ô¹ÏÍø is proud to announce that our research papers have been accepted at (CVPR). As the most prestigious event in the field of computer vision, CVPR is the ultimate platform for showcasing ground-breaking research and innovations that have the potential to redefine the technological landscape.

Research papers accepted at CVPR'24:

NRDF: Neural Riemannian Distance Fields for Learning Articulated Pose Priors Yannan He (University of Tübingen) · Garvita Tiwari (University of Tuebingen and MPI-Saarbrucken) · Tolga Birdal (51³Ô¹ÏÍø) · Jan Lenssen (Saarland Informatics Campus, Max-Planck Institute) · Gerard Pons-Moll (University of Tübingen)

HyperSDFusion: Bridging Hierarchical Structures in Language and Geometry for Enhanced 3D Text2Shape Generation · Zhiying Leng (Technical University of Munich) · Tolga Birdal (51³Ô¹ÏÍø) · Xiaohui Liang (Zhongguancun Laboratory) · Federico Tombari (Google, TUM)

Fun with Flags: Robust Principal Directions via Flag Manifolds · Tolga Birdal (51³Ô¹ÏÍø) · Nathan Mankovich (University of Valencia)

 Probabilistic Sampling of Balanced K-Means using Adiabatic Quantum Computing · Jan-Nico Zaech (ETH Zürich) · Martin Danelljan (ETH Zurich) · Tolga Birdal (51³Ô¹ÏÍø) · Luc Van Gool (ETH Zurich)

Gaussian Splatting SLAM · Hidenobu Matsuki (51³Ô¹ÏÍø) · Riku Murai (51³Ô¹ÏÍø) · Paul Kelly (51³Ô¹ÏÍø) · Andrew J. Davison (51³Ô¹ÏÍø)

SuperPrimitive: Scene Reconstruction at a Primitive Level · Kirill Mazur (51³Ô¹ÏÍø) · Gwangbin Bae (51³Ô¹ÏÍø) · Andrew J. Davison (51³Ô¹ÏÍø)

EscherNet: A Generative Model for Scalable View Synthesis · Xin Kong (51³Ô¹ÏÍø) · Shikun Liu (51³Ô¹ÏÍø) · Xiaoyang Lyu (University of Hong Kong) · Marwan Taher (51³Ô¹ÏÍø) · Xiaojuan Qi (University of Hong Kong) · Andrew J. Davison (51³Ô¹ÏÍø)

Rethinking Inductive Biases for Surface Normal Estimation · Gwangbin Bae (51³Ô¹ÏÍø) · Andrew J. Davison (51³Ô¹ÏÍø)

Design2Cloth: 3D Cloth Generation from 2D Masks · Jiali Zheng (51³Ô¹ÏÍø) · Rolandos Alexandros Potamias (51³Ô¹ÏÍø) · Stefanos Zafeiriou (51³Ô¹ÏÍø)

Locally Adaptive Neural 3D Morphable Models · Michail Tarasiou (51³Ô¹ÏÍø) · Rolandos Alexandros Potamias (51³Ô¹ÏÍø) · Eimear O' Sullivan (Huawei Technologies Ltd.) · Stylianos Ploumpis (51³Ô¹ÏÍø) · Stefanos Zafeiriou (51³Ô¹ÏÍø)

Neural Sign Actors: A diffusion model for 3D sign language production from text · Vasileios Baltatzis (None) · Rolandos Alexandros Potamias (51³Ô¹ÏÍø) · Evangelos Ververas (Huawei Technologies Ltd.) · Guanxiong Sun (Huawei Technologies Ltd.) · Jiankang Deng (51³Ô¹ÏÍø & Huawei UKRD) · Stefanos Zafeiriou (51³Ô¹ÏÍø)

G-FARS: Gradient-Field-based Auto-Regressive Sampling for 3D Part Grouping · Junfeng Cheng (51³Ô¹ÏÍø) · Tania Stathaki (51³Ô¹ÏÍø)

CVPR provides an unparalleled forum for researchers from around the globe to share their insights and discoveries, fostering collaboration and sparking innovation. Dr. Birdal’s presence at this esteemed conference places him among the world's leading visionaries in computer vision, highlighting the department's role as a hub of pioneering research and technological progress.

Research papers accepted at ICLR'24:

Research papers were also accepted at (ICLR) 2024 which is is a machine learning conference.

Variational Inference for SDEs Driven by Fractional Noise (spotlight) · Rembert Daems · Manfred Opper · Guillaume Crevecoeur · Tolga Birdal (51³Ô¹ÏÍø)

C-TPT: Calibrated Test-Time Prompt Tuning for Vision-Language Models via Text Feature Dispersion · Hee Suk Yoon · Eunseop Yoon · Joshua Tian Jin Tee · Mark A. Hasegawa-Johnson · Yingzhen Li (51³Ô¹ÏÍø) · Chang D. Yoo

Grounded Object-Centric Learning · Avinash Kori · Francesco Locatello · Fabio De Sousa Ribeiro · Francesca Toni · Ben Glocker (51³Ô¹ÏÍø)

Post-hoc Bias Scoring Is Optimal For Fair Classification. ICLR 2024 (spotlight) · Wenlong Chen* (51³Ô¹ÏÍø) · Yegor Klochkov* · Yang Liu

The Department of Computing is committed to supporting our researchers and academics in their endeavors to explore uncharted territories and solve complex challenges. The success is a testament to the vibrant research culture and intellectual rigor that define our community.

We invite the 51³Ô¹ÏÍø College community and the wider public to join us in congratulating our researchers on this outstanding achievement. their work not only advances the field of computer vision but also inspires the next generation of scientists and engineers to pursue their own path of innovation and discovery.