Vilio: State-of-the-art Visio-Linguistic Models applied to Hateful Memes

Autor:	Muennighoff, Niklas
Rok vydání:	2020
Předmět:	Computer Science - Artificial Intelligence Computer Science - Computation and Language Computer Science - Computer Vision and Pattern Recognition
Druh dokumentu:	Working Paper
Popis:	This work presents Vilio, an implementation of state-of-the-art visio-linguistic models and their application to the Hateful Memes Dataset. The implemented models have been fitted into a uniform code-base and altered to yield better performance. The goal of Vilio is to provide a user-friendly starting point for any visio-linguistic problem. An ensemble of 5 different V+L models implemented in Vilio achieves 2nd place in the Hateful Memes Challenge out of 3,300 participants. The code is available at https://github.com/Muennighoff/vilio. Comment: Presented at NIPS 2020
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2012.07788 Zobrazit plný text záznamu View this record from Arxiv