Evaluation of Novel AI Architectures for Uncertainty Estimation
Deep learning (DL) has advanced computer vision, delivering impressive performance on intricate visual tasks. Yet, the need for accurate uncertainty estimations, particularly for out-of-distribution (OOD) inputs, persists. Our research evaluates uncertainty in Convolutional Neural Networks (CNN) an...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Universidad Autónoma de Bucaramanga
2024-12-01
|
| Series: | Revista Colombiana de Computación |
| Subjects: | |
| Online Access: | https://revistas.unab.edu.co/index.php/rcc/article/view/5274 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832584085532311552 |
|---|---|
| author | Erik Pautsch John Li Silvio Rizzi George K. Thiruvathukal Maria Pantoja |
| author_facet | Erik Pautsch John Li Silvio Rizzi George K. Thiruvathukal Maria Pantoja |
| author_sort | Erik Pautsch |
| collection | DOAJ |
| description |
Deep learning (DL) has advanced computer vision, delivering impressive performance on intricate visual tasks. Yet, the need for accurate uncertainty estimations, particularly for out-of-distribution (OOD) inputs, persists. Our research evaluates uncertainty in Convolutional Neural Networks (CNN) and Vision Transformers (ViT) using the MNIST and ImageNet-1K datasets. Using High-Performance (HPC) platforms, including the traditional Polaris supercomputer and AI accelerators like Cerebras CS-2 and SambaNova DataScale, we assessed the computational merits and bottlenecks of each platform. This paper delineates key considerations for using HPC in uncertainty estimations in DL, offering insights that guide the integration of algorithms and hardware for robust DL applications, especially in computer vision.
|
| format | Article |
| id | doaj-art-a6cd17dd2559469aaaaddc342b38a65a |
| institution | Kabale University |
| issn | 1657-2831 2539-2115 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Universidad Autónoma de Bucaramanga |
| record_format | Article |
| series | Revista Colombiana de Computación |
| spelling | doaj-art-a6cd17dd2559469aaaaddc342b38a65a2025-01-27T19:31:14ZengUniversidad Autónoma de BucaramangaRevista Colombiana de Computación1657-28312539-21152024-12-0125210.29375/25392115.5274Evaluation of Novel AI Architectures for Uncertainty EstimationErik Pautsch0John Li1Silvio Rizzi2George K. Thiruvathukal3Maria Pantoja4https://orcid.org/0000-0002-1942-9769Loyola University ChicagoUniversity of California San DiegoArgonne National LaboratoryLoyola University ChicagoCalifornia Polytechnic State University Deep learning (DL) has advanced computer vision, delivering impressive performance on intricate visual tasks. Yet, the need for accurate uncertainty estimations, particularly for out-of-distribution (OOD) inputs, persists. Our research evaluates uncertainty in Convolutional Neural Networks (CNN) and Vision Transformers (ViT) using the MNIST and ImageNet-1K datasets. Using High-Performance (HPC) platforms, including the traditional Polaris supercomputer and AI accelerators like Cerebras CS-2 and SambaNova DataScale, we assessed the computational merits and bottlenecks of each platform. This paper delineates key considerations for using HPC in uncertainty estimations in DL, offering insights that guide the integration of algorithms and hardware for robust DL applications, especially in computer vision. https://revistas.unab.edu.co/index.php/rcc/article/view/5274UncertaintyDeep LearningEnsemblesEvidential LearningArtificial intelligence |
| spellingShingle | Erik Pautsch John Li Silvio Rizzi George K. Thiruvathukal Maria Pantoja Evaluation of Novel AI Architectures for Uncertainty Estimation Revista Colombiana de Computación Uncertainty Deep Learning Ensembles Evidential Learning Artificial intelligence |
| title | Evaluation of Novel AI Architectures for Uncertainty Estimation |
| title_full | Evaluation of Novel AI Architectures for Uncertainty Estimation |
| title_fullStr | Evaluation of Novel AI Architectures for Uncertainty Estimation |
| title_full_unstemmed | Evaluation of Novel AI Architectures for Uncertainty Estimation |
| title_short | Evaluation of Novel AI Architectures for Uncertainty Estimation |
| title_sort | evaluation of novel ai architectures for uncertainty estimation |
| topic | Uncertainty Deep Learning Ensembles Evidential Learning Artificial intelligence |
| url | https://revistas.unab.edu.co/index.php/rcc/article/view/5274 |
| work_keys_str_mv | AT erikpautsch evaluationofnovelaiarchitecturesforuncertaintyestimation AT johnli evaluationofnovelaiarchitecturesforuncertaintyestimation AT silviorizzi evaluationofnovelaiarchitecturesforuncertaintyestimation AT georgekthiruvathukal evaluationofnovelaiarchitecturesforuncertaintyestimation AT mariapantoja evaluationofnovelaiarchitecturesforuncertaintyestimation |