MViTv2: Improved Multiscale Vision Transformers for Classification and Detection

Code

• github.com/facebookresearch/SlowFast
  OfficialIn paperpytorch★ 7,320
• github.com/facebookresearch/mvit
  Officialpytorch★ 453
• github.com/facebookresearch/detectron2/tree/main/projects/MViTv2
  Officialpytorch★ 0
• github.com/rwightman/pytorch-image-models
  pytorch★ 36,538
• github.com/open-mmlab/mmclassification
  pytorch★ 3,839
• github.com/JunweiLiang/aicity_action
  pytorch★ 28
• github.com/3dperceptionlab/visual-wetlandbirds
  pytorch★ 9
• github.com/rajatmodi62/occludedactionbenchmark
  pytorch★ 9
• gitlab.com/birder/birder
  pytorch★ 0

Abstract

In this paper, we study Multiscale Vision Transformers (MViTv2) as a unified architecture for image and video classification, as well as object detection. We present an improved version of MViT that incorporates decomposed relative positional embeddings and residual pooling connections. We instantiate this architecture in five sizes and evaluate it for ImageNet classification, COCO detection and Kinetics video recognition where it outperforms prior work. We further compare MViTv2s' pooling attention to window attention mechanisms where it outperforms the latter in accuracy/compute. Without bells-and-whistles, MViTv2 has state-of-the-art performance in 3 domains: 88.8% accuracy on ImageNet classification, 58.7 boxAP on COCO object detection as well as 86.1% on Kinetics-400 video classification. Code and models are available at https://github.com/facebookresearch/mvit.

Tasks

Benchmark Results

Reproductions