Pregled bibliografske jedinice broj: 1146433
BricksND - Benchmark for disentangled concept learning
BricksND - Benchmark for disentangled concept learning // 5th International Workshop on Data Science (IWDS 2020)
Zagreb, Hrvatska, 2020. (predavanje, nije recenziran, pp prezentacija, znanstveni)
CROSBI ID: 1146433 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
BricksND - Benchmark for disentangled concept
learning
Autori
Grubišić, Ivan ; Lipić, Tomislav ; Mesić, Benjamin ; Šmuc, Tomislav
Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, pp prezentacija, znanstveni
Skup
5th International Workshop on Data Science (IWDS 2020)
Mjesto i datum
Zagreb, Hrvatska, 24.11.2020
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Nije recenziran
Ključne riječi
disentangled representations ; benchmark datasets ; concept-based representation learning ; explainable artificial intelligence
Sažetak
This presentation talks about a novel benchmark for disentangled concept learning, called “BricksND”. The benchmark consists of synthetically generated brick scenes in 1D, 2D and 3D spaces. This benchmark is designed to address the challenges of machine learning, especially disentanglement representations and concept learning.
Izvorni jezik
Engleski
Znanstvena područja
Računarstvo
POVEZANOST RADA
Ustanove:
Institut "Ruđer Bošković", Zagreb
