Publications by Claudiu-Cristian Musat

Benefiting From Bicubically Down-Sampled Images for Learning Real-World Image Super-Resolution

Jean-Philippe Thiran, Thomas Yu, Seyedbehzad Bozorgtabar, Claudiu-Cristian Musat, Hazim Kemal Ekenel, Mohammad Saeed Rad

Super-resolution (SR) has traditionally been based on pairs of high-resolution images (HR) and their low-resolution (LR) counterparts obtained artificially with bicubic downsampling. However, in real-world SR, there is a large variety of realistic image de ...

2021

Interacting with Explanations through Critiquing

Boi Faltings, Claudiu-Cristian Musat, Diego Matteo Antognini

Using personalized explanations to support recommendations has been shown to increase trust and perceived quality. However, to actually obtain better recommendations, there needs to be a means for users to modify the recommendation criteria by interacting ...

2021

Recommending Burgers based on Pizza Preferences: Addressing Data Sparsity with a Product of Experts

Claudiu-Cristian Musat, Diego Matteo Antognini, Martin Milenkoski

In this paper, we describe a method to tackle data sparsity and create recommendations in domains with limited knowledge about user preferences. We expand the variational autoencoder collaborative filtering from a single-domain to a multi-domain setting. T ...

2021

Momentum-based Gradient Methods in Multi-Objective Recommendation

Boi Faltings, Claudiu-Cristian Musat, Diego Matteo Antognini

Multi-objective gradient methods are becoming the standard for solving multi-objective problems. Among others, they show promising results in developing multi-objective recommender systems with both correlated and conflicting objectives. Classic multi-grad ...

2021

Modeling Online Behavior in Recommender Systems: The Importance of Temporal Context

Boi Faltings, Claudiu-Cristian Musat, Diego Matteo Antognini

Simulating online recommender system performance is no- toriously difficult and the discrepancy between the online and offline behaviors is typically not accounted for in offline evaluations. Recommender systems research tends to evalu- ate model performan ...

2021

Multi-Dimensional Explanation of Target Variables from Documents

Boi Faltings, Claudiu-Cristian Musat, Diego Matteo Antognini

Automated predictions require explanations to be interpretable by humans. Past work used attention and rationale mechanisms to find words that predict the target variable of a document. Often though, they result in a tradeoff between noisy explanations or ...

ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE2021

Addressing fairness in classification with a model-agnostic multi-objective algorithm

Boi Faltings, Claudiu-Cristian Musat, Diego Matteo Antognini

The goal of fairness in classification is to learn a classifier that does not discriminate against groups of individuals based on sensitive attributes, such as race and gender. One approach to designing fair algorithms is to use relaxations of fairness not ...

2021

Benefiting from Multitask Learning to Improve Single Image Super-Resolution

Jean-Philippe Thiran, Mohammad Saeed Rad, Claudiu-Cristian Musat, Hazim Kemal Ekenel

Despite significant progress toward super resolving more realistic images by deeper convolutional neural networks (CNNs), reconstructing fine and natural textures still remains a challenging problem. Recent works on single image super resolution (SISR) are ...

2020

Evaluating the search phase of neural architecture search

Mathieu Salzmann, Martin Jaggi, Claudiu-Cristian Musat, Kaicheng Yu

Neural Architecture Search (NAS) aims to facilitate the design of deep networks fornew tasks. Existing techniques rely on two stages: searching over the architecture space and validating the best architecture. NAS algorithms are currently compared solely b ...

2020

Overcoming Multi-model Forgetting

Mathieu Salzmann, Anthony Christopher Davison, Martin Jaggi, Claudiu-Cristian Musat, Kaicheng Yu, Yassine Benyahia

We identify a phenomenon, which we refer to as multi-model forgetting, that occurs when sequentially training multiple deep networks with partially-shared parameters; the performance of previously-trained models degrades as one optimizes a subsequent one, ...

JMLR2019