NeurIPS'22 Cross-Domain MetaDL Challenge: Results and lessons learned - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

NeurIPS'22 Cross-Domain MetaDL Challenge: Results and lessons learned

NeurIPS'22 Cross-Domain MetaDL Défi: Résultats et enseignements

Mahbubul Alam
  • Fonction : Auteur
Ahmed Farahat
  • Fonction : Auteur
Dipanjan Ghosh
  • Fonction : Auteur
Chetan Gupta
  • Fonction : Auteur
Xian Yeow Lee
  • Fonction : Auteur
Manh Hung Nguyen
  • Fonction : Auteur

Résumé

Deep neural networks have demonstrated the ability to outperform humans in multiple tasks, but they often require substantial amounts of data and computational resources. These resources may be limited in certain fields. Meta-learning seeks to overcome these challenges by utilizing past task experiences to efficiently solve new tasks, achieving better performance with limited training data and modest computational resources. To further advance the ChaLearn MetaDL competition series, we organized the Cross-Domain MetaDL Challenge for NeurIPS'22. This challenge aimed to solve "any-way" and "any-shot" tasks from 10 domains through cross-domain meta-learning. In this paper, authored collaboratively by the competition organizers, top-ranked participants, and external collaborators, we describe the technical aspects of the competition, baseline methods, and top-ranked approaches that have been open-sourced. Additionally, we provide a detailed analysis of the competition results. Lessons learned from this competition include the critical role of pre-trained backbones, the necessity of preventing overfitting, and the significance of using data augmentation or domain adaptation techniques in conjunction with extra optimizations to improve performance.
Fichier principal
Vignette du fichier
carrion-ojeda22a.pdf (560.2 Ko) Télécharger le fichier
Origine Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

hal-04198139 , version 1 (08-09-2023)

Licence

Identifiants

Citer

Dustin Carrión-Ojeda, Mahbubul Alam, Sergio Escalera, Ahmed Farahat, Dipanjan Ghosh, et al.. NeurIPS'22 Cross-Domain MetaDL Challenge: Results and lessons learned. Proceedings of the NeurIPS 2022 Competitions Track, Nov 2022, New Orleans (Louisiana), United States. pp.50-72. ⟨hal-04198139⟩
63 Consultations
45 Téléchargements

Altmetric

Partager

Gmail Mastodon Facebook X LinkedIn More