TY  - JOUR
AB  - As the field of deep learning steadily transitions from the realm of academic research to practical application, the significance of self-supervised pretraining methods has become increasingly prominent. These methods, particularly in the image domain, offer a compelling strategy to effectively utilize the abundance of unlabeled image data, thereby enhancing downstream tasks’ performance. In this paper, we propose Spatial Reasoning, a novel auxiliary pretraining method that takes advantage of a more flexible formulation of contrastive learning by introducing spatial reasoning as an auxiliary task for discriminative self-supervised methods. Spatial Reasoning works by having the network predict the relative distances between sampled non-overlapping patches. We argue that this forces the network to learn more detailed and intricate internal representations of the objects and the relationships between their constituting parts. Our experiments demonstrate substantial improvement in downstream performance in linear evaluation compared to similar work and provide directions for further research into spatial reasoning.
AU  - Albert, T.*
AU  - Eskofier, B.M.
AU  - Zanca, D.*
C1  - 70594
C2  - 55588
CY  - One New York Plaza, Suite 4600, New York, Ny, United States
TI  - From patches to objects: Exploiting spatial reasoning for better visual representations.
JO  - SN appl. sci.
VL  - 6
IS  - 5
PB  - Springer
PY  - 2024
SN  - 2523-3963
ER  - 

TY  - JOUR
AB  - The aim of this work is the synthesis and characterization of C, N, S–Fe multidoped TiO2
       photocatalysts. The nanoparticles were prepared by sol–gel method and 
      investigated by X-ray diffraction (XRD), transmission electron 
      microscopy, UV–visible diffuse reflectance spectroscopy (DRS) and N2
       adsorption–desorption isotherm (BET). Photocatalytic activity of the 
      photocatalysts was studied by degrading methyl orange (MO) under UV and 
      visible light irradiations. XRD analysis illustrated that all 
      synthesized samples present the anatase phase. CNS–0.3Fe-doped TiO2 exhibited the highest surface area of 891 m2/g.
       DRS showed that Fe doping had efficiently narrowed the band gap to 
      2.74 eV and broaden the visible-light response region. The synthesized 
      nanomaterials exhibited high photodegradation ability in the 
      decomposition of MO than commercial pure TiO2 nanoparticles Degussa P25.
AU  - Ellouzi, I.*
AU  - El Hajjaji, S.*
AU  - Harir, M.
AU  - Schmitt-Kopplin, P.
AU  - Robert, D.*
AU  - Laanab, L.*
C1  - 57824
C2  - 47933
TI  - Synergistic effects of C, N, S, Fe-multidoped TiO2 for photocatalytic degradation of methyl orange dye under UV and visible light irradiations.
JO  - SN appl. sci.
VL  - 1
IS  - 8
PY  - 2019
SN  - 2523-3963
ER  -