Which Transformer to Favor: A Comparative Analysis of Efficiency in Vision Transformers

arXiv:2308.09372v2 Announce Type: replace-cross
Abstract: Transformers come with a high computational cost, yet their effectiveness in addressing problems in language and vision has sparked extensive research aimed at enhancing their efficiency. However, diverse experimental conditions, spanning multiple input domains, prevent a fair comparison based solely on reported results, posing challenges for model selection. To address this gap in comparability, we design a comprehensive benchmark of more than 30 models for image classification, evaluating key efficiency aspects, including accuracy, speed, and …

abstract analysis arxiv comparative analysis comparison computational cost cs.ai cs.cv cs.lg diverse domains efficiency experimental fair however language multiple research results transformer transformers type vision vision transformers