Vision reading group ‒ VITA ‐ EPFL

Welcome to our page!

We host (virtual) seminars on machine learning and computer vision topics. Our targets are research papers published at renowned conferences in the field (CVPR, ICCV, NeurIPS, ICML, etc.). Every week, authors of inspiring papers are invited to present their work.

Below, you’ll find the list of previous talks. For the latest updated, please subscribe to our Youtube channel.

Fall 2020

Name & Lab	Date & Time	Papers	Video Recording
Pim de Haan (University of Amsterdam)	17 Sep 3 pm	Causal Confusion in Imitation Learning (NeurIPS 2019, oral)
Yonglong Tian (MIT)	24 Sep 5 pm	What Makes for Good Views for Contrastive Learning (NeurIPS 2020)
Zuxuan Wu (University of Maryland / Facebook AI)	1 Oct 5 pm	Making an Invisibility Cloak: Real World Adversarial Attacks on Object Detectors (ECCV 2020)	NA
Francesco Locatello (ETH)	8 Oct 3 pm	Object-Centric Learning with Slot Attention (NeurIPS 2020)
Sergey Zakharov (TUM / TRI)	15 Oct 3 pm	Autolabeling 3D Objects with Differentiable Rendering of 3D Shapes Priors (CVPR 2020, oral)
Sergio Casas (U. of Toronto / Uber ATG)	22 Oct 5 pm	Spatially-Aware Graph Neural Networks for Relational Behavior Forecasting from Sensor Data (ICRA 2020) The Importance of Prior Knowledge in Precise Multimodal Prediction (IROS 2020) Implicit Latent Variable Model for Scene-Consistent Motion Forecasting (ECCV 2020)
Marc Habermann (MPI)	29 Oct 5 pm	DeepCap: Monocular Human Performance Capture Using Weak Supervision (CVPR 2020, Oral)
	CVPR Break
Weili Nie (Rice University)	26 Nov 5 pm	Semi-Supervised StyleGAN for Disentanglement Learning (ICML 2020)
Matthew Tancik (UC Berkeley)	3 Dec 5 pm	Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains (NeurIPS 2020)
Annie Xie (Stanford)	10 Dec 5 pm	Learning Latent Representations to Influence Multi-Agent Interaction (COrL 2020, best paper award)
Adam Gleave (UC Berkeley)	17 Dec 5 pm	Adversarial Policies: Attacking Deep Reinforcement Learning (ICLR 2020)

For inquiries, please contact Lorenzo Bertoni or Saeed Saadatnejad.