Emergent multi-agent communication in the deep learning era
The ability to cooperate through language is a defining feature of humans. As the
perceptual, motory and planning capabilities of deep artificial networks increase …
perceptual, motory and planning capabilities of deep artificial networks increase …
[HTML][HTML] Learning disentangled representations in the imaging domain
Disentangled representation learning has been proposed as an approach to learning
general representations even in the absence of, or with limited, supervision. A good general …
general representations even in the absence of, or with limited, supervision. A good general …
Toward causal representation learning
The two fields of machine learning and graphical causality arose and are developed
separately. However, there is, now, cross-pollination and increasing interest in both fields to …
separately. However, there is, now, cross-pollination and increasing interest in both fields to …
Disentangled representation learning
Disentangled Representation Learning (DRL) aims to learn a model capable of identifying
and disentangling the underlying factors hidden in the observable data in representation …
and disentangling the underlying factors hidden in the observable data in representation …
Self-supervised learning with data augmentations provably isolates content from style
Self-supervised representation learning has shown remarkable success in a number of
domains. A common practice is to perform data augmentation via hand-crafted …
domains. A common practice is to perform data augmentation via hand-crafted …
Causal intervention for weakly-supervised semantic segmentation
We present a causal inference framework to improve Weakly-Supervised Semantic
Segmentation (WSSS). Specifically, we aim to generate better pixel-level pseudo-masks by …
Segmentation (WSSS). Specifically, we aim to generate better pixel-level pseudo-masks by …
Causality for machine learning
B Schölkopf - Probabilistic and causal inference: The works of Judea …, 2022 - dl.acm.org
The machine learning community's interest in causality has significantly increased in recent
years. My understanding of causality has been shaped by Judea Pearl and a number of …
years. My understanding of causality has been shaped by Judea Pearl and a number of …
Counterfactual zero-shot and open-set visual recognition
We present a novel counterfactual framework for both Zero-Shot Learning (ZSL) and Open-
Set Recognition (OSR), whose common challenge is generalizing to the unseen-classes by …
Set Recognition (OSR), whose common challenge is generalizing to the unseen-classes by …
Interventional few-shot learning
We uncover an ever-overlooked deficiency in the prevailing Few-Shot Learning (FSL)
methods: the pre-trained knowledge is indeed a confounder that limits the performance. This …
methods: the pre-trained knowledge is indeed a confounder that limits the performance. This …
Disentangling user interest and conformity for recommendation with causal embedding
Recommendation models are usually trained on observational interaction data. However,
observational interaction data could result from users' conformity towards popular items …
observational interaction data could result from users' conformity towards popular items …