In object detection and many other computer vision benchmarks, the image resolutions as well as the aspect ratios are usually not fixed as the image classification task. For the first layer, the PE is interpolated following ViT. In a word, Type-I uses more PEs and Type-II uses larger PE. In our paper, small-and base-sized models use this setting. The detailed configurations are given in Tab. 1. PE-cls to PE-det Rand.

In object detection and many other computer vision benchmarks, the image resolutions as well as the aspect ratios are usually not fixed as the image classification task. For the first layer, the PE is interpolated following ViT. In a word, Type-I uses more PEs and Type-II uses larger PE. In our paper, small-and base-sized models use this setting. The detailed configurations are given in Tab. 1. PE-cls to PE-det Rand.

Can Transformer perform $2\mathrm{D}$ object-level recognition from a pure sequence-to-sequence perspective with minimal knowledge about the $2\mathrm{D}$ spatial structure? To answer this question, we present You Only Look at One Sequence (YOLOS), a series of object detection models based on the na\"ive Vision Transformer with the fewest possible modifications as well as inductive biases. We find that YOLOS pre-trained on the mid-sized ImageNet-$1k$ dataset only can already achieve competitive object detection performance on COCO, \textit{e.g.}, YOLOS-Base directly adopted from BERT-Base can achieve $42.0$ box AP. We also discuss the impacts as well as limitations of current pre-train schemes and model scaling strategies for Transformer in vision through object detection. Code and model weights are available at \url{https://github.com/hustvl/YOLOS}.