Zobrazeno 1 - 1
of 1
pro vyhledávání: '"Barseghyan, Alvard"'
Autor:
Vanyan, Ani, Barseghyan, Alvard, Tamazyan, Hakob, Huroyan, Vahan, Khachatrian, Hrant, Danelljan, Martin
In this paper, we present a comparative analysis of various self-supervised Vision Transformers (ViTs), focusing on their local representative power. Inspired by large language models, we examine the abilities of ViTs to perform various computer visi
Externí odkaz:
http://arxiv.org/abs/2401.00463
