MViTv2: Improved Multiscale Vision Transformers for Classification and Detection
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.
Predict image
from mmpretrain import inference_model
predict = inference_model('mvitv2-tiny_3rdparty_in1k', 'demo/bird.JPEG')
print(predict['pred_class'])
print(predict['pred_score'])Use the model
import torch
from mmpretrain import get_model
model = get_model('mvitv2-tiny_3rdparty_in1k', pretrained=True)
inputs = torch.rand(1, 3, 224, 224)
out = model(inputs)
print(type(out))
# To extract features.
feats = model.extract_feat(inputs)
print(type(feats))Test Command
Prepare your dataset according to the docs.
Test:
python tools/test.py configs/mvit/mvitv2-tiny_8xb256_in1k.py https://download.openmmlab.com/mmclassification/v0/mvit/mvitv2-tiny_3rdparty_in1k_20220722-db7beeef.pth| Model | Pretrain | Params (M) | Flops (G) | Top-1 (%) | Top-5 (%) | Config | Download |
|---|---|---|---|---|---|---|---|
mvitv2-tiny_3rdparty_in1k* |
From scratch | 24.17 | 4.70 | 82.33 | 96.15 | config | model |
mvitv2-small_3rdparty_in1k* |
From scratch | 34.87 | 7.00 | 83.63 | 96.51 | config | model |
mvitv2-base_3rdparty_in1k* |
From scratch | 51.47 | 10.16 | 84.34 | 96.86 | config | model |
mvitv2-large_3rdparty_in1k* |
From scratch | 217.99 | 43.87 | 85.25 | 97.14 | config | model |
Models with * are converted from the official repo. The config files of these models are only for inference. We haven't reproduce the training results.
@inproceedings{li2021improved,
title={MViTv2: Improved multiscale vision transformers for classification and detection},
author={Li, Yanghao and Wu, Chao-Yuan and Fan, Haoqi and Mangalam, Karttikeya and Xiong, Bo and Malik, Jitendra and Feichtenhofer, Christoph},
booktitle={CVPR},
year={2022}
}