Vision Transformers in 2022: An Update on Tiny ImageNet
Authors:
Ethan Huynh
Abstract:
The recent advances in image transformers have shown impressive results and have largely closed the gap between traditional CNN architectures. The standard procedure is to train on large datasets like ImageNet-21k and then finetune on ImageNet-1k. After finetuning, researches will often consider the transfer learning performance on smaller datasets such as CIFAR-10/100 but have left out Tiny Image…
▽ More
The recent advances in image transformers have shown impressive results and have largely closed the gap between traditional CNN architectures. The standard procedure is to train on large datasets like ImageNet-21k and then finetune on ImageNet-1k. After finetuning, researches will often consider the transfer learning performance on smaller datasets such as CIFAR-10/100 but have left out Tiny ImageNet. This paper offers an update on vision transformers' performance on Tiny ImageNet. I include Vision Transformer (ViT) , Data Efficient Image Transformer (DeiT), Class Attention in Image Transformer (CaiT), and Swin Transformers. In addition, Swin Transformers beats the current state-of-the-art result with a validation accuracy of 91.35%. Code is available here: https://github.com/ehuynh1106/TinyImageNet-Transformers
△ Less
Submitted 21 May, 2022;
originally announced May 2022.
Designing Narrative-Focused Role-Playing Games for Visualization Literacy in Young Children
Authors:
Elaine Huynh,
Angela Nyhout,
Patricia Ganea,
Fanny Chevalier
Abstract:
Building on game design and education research, this paper introduces narrative-focused role-playing games as a way to promote visualization literacy in young children. Visualization literacy skills are vital in understanding the world around us and constructing meaningful visualizations, yet, how to better develop these skills at an early age remains largely overlooked and understudied. Only rece…
▽ More
Building on game design and education research, this paper introduces narrative-focused role-playing games as a way to promote visualization literacy in young children. Visualization literacy skills are vital in understanding the world around us and constructing meaningful visualizations, yet, how to better develop these skills at an early age remains largely overlooked and understudied. Only recently has the visualization community started to fill this gap, resulting in preliminary studies and development of educational tools for use in early education. We add to these efforts through the exploration of gamification to support learning, and identify an opportunity to apply role-playing game-based designs by leveraging the presence of narratives in data-related problems involving visualizations. We study the effects of including narrative elements on learning through a technology probe, grounded in a set of design considerations stemming from visualization, game design, and education science. We create two versions of a game -- one with narrative elements and one without -- and evaluate our instances on 33 child participants between 11- to 13-years old using a between-subjects study design. Despite participants requiring double the amount of time to complete their game due to additional elements, the inclusion of such elements were found to improve engagement without sacrificing learning; our results indicate no significant differences in development of graph-reading skills, but significant differences in engagement and overall enjoyment of the game. We report observations and qualitative feedback collected, and note areas for improvement and room for future wook.
△ Less
Submitted 31 August, 2020;
originally announced August 2020.
