Talks

The program is still subject to minor modifications.

Morning

9:00 Xavier Vilajosana (UOC) - Welcome

9:05 Dimosthenis Karatzas (UAB - CVC) - Presentation of the ELLIS Unit Barcelona.


9:10 Keynote:

Local Chair: Cristian Canton (Meta)

Fernando de la Torre (CMU), Human Sensing 


10:00 Session 1

Local Chair: Petia Radeva (UB)

Session Chair: Aleix Martínez (Amazon)

German Barquero (UB - CVC) - BeLFusion: Latent Diffusion for Behavior-Driven Human Motion Prediction (ICCV 2023)

Wentong Liao (Amazon) - Network-free, unsupervised semantic segmentation with synthetic images (CVPR 2023)

Margarita Geleta (Stanford) - Adversarial Learning for Feature Shift Detection and Correction (NeurIPS 2023)

Xavi Suau (Apple) - DUET: 2D Structured and Approximately Equivariant Representations (ICML 2023)



12:00 Session 2

Local Chair: Coloma Ballester (UPF)

Session Chair: Antonio Torralba (MIT)

Ginger Delmas (Naver - IRI) - PoseFix: Correcting 3D Human Poses with Natural Language (ICCV 2023)

Lluís Castrejon (Google) - Encyclopedic VQA: Visual questions about detailed properties of fine-grained categories (ICCV 2023)

Dídac Surís (Columbia) - ViperGPT: Visual Inference via Python Execution for Reasoning (ICCV 2023)

Victor Campos (Deepmind) - Human-level Atari 200x faster (ICLR 2023)

Maria Bauza (Deepmind) - RoboCat: A Self-Improving Foundation Agent for Robotic Manipulation (follow up of IJRR 2022)

Ferran Alet (Deepmind) - Leveraging GNNs for skillfull weather forecasting (Science 2023)

Afternoon

15:00 Session 3

Local Chair: Xavier Giró-i-Nieto (UPC-Amazon)

Session Chair: Marta R. Costa-jussà (Meta) 

Adrià Mallol-Ragolta (Augsburg) - The MASCFLICHT Corpus: Face Mask Type and Coverage Area Recognition from Speech (Interspeech 2023)

Ioannis Tsiamas (UPC) - SegAugment: Maximizing the Utility of Speech Translation Data with Segmentation-based Augmentations  (EMNLP 2023)

Gerard Sant (UPC - BSC) - Analysis of Acoustic information in End-to-End Spoken Language Translation (Interspeech 2023)

Pedro Ramoneda (UPF) - Predicting performance difficulty from piano sheet music images (ISMIR 2023)

Joan Serrà (Dolby) - Mono-to-stereo through parametric stereo generation (ISMIR 2023)

Juan Montesinos (UPF) - Speech Inpainting: Context-based Speech Synthesis Guided by Video (Interspeech 2023)

Marco Baroni (UPF) - Can discrete information extraction prompts generalize across language models? (ICLR 2023)



17:30 Session 4 

Local Chair: Àgata Lapedriza  (UOC-NEU) 

Session Chair: Ricardo Baeza-Yates (UPF-NEU-UChile) 

Javier Ferrando (UPC) - Explaining How Transformers Use Context to Build Predictions (ACL 2023)

Dipam Goswami (CVC-UAB) - FeCAM: Exploiting the Heterogeneity of Class Distributions in Exemplar-Free Continual Learning (NeurIPS 2023)

Yannis Kalantidis (Naver Labs) - Fake it till you make it: Learning transferable representations from synthetic ImageNet clones (CVPR 2023)

Imanol G. Estepa (UB-CVC) - All4One: Symbiotic Neighbour Contrastive Learning via Self-Attention and Redundancy Reduction (ICCV 2023)

Belen Alastruey (Amazon) - Multi-view frequency-attention alternative to CNN frontends for automatic speech recognition (Interspeech 2023)

Guillem Simeon (UPF) - TensorNet: Cartesian Tensor Representations for Efficient Learning of Molecular Potentials (NeurIPS 2023)


18:50 Marta Aymerich & Carles Ventura (UOC) - Closing