SparK

Designing BERT for Convolutional Networks: Sparse and Hierarchical Masked Modeling

Abstract

We identify and overcome two key obstacles in extending the success of BERT-style pre-training, or the masked image modeling, to convolutional networks (convnets): (i) convolution operation cannot handle irregular, random-masked input images; (ii) the single-scale nature of BERT pre-training is inconsistent with convnet’s hierarchical structure. For (i), we treat unmasked pixels as sparse voxels of 3D point clouds and use sparse convolution to encode. This is the first use of sparse convolution for 2D masked modeling. For (ii), we develop a hierarchical decoder to reconstruct images from multi-scale encoded features. Our method called Sparse masKed modeling (SparK) is general: it can be used directly on any convolutional model without backbone modifications. We validate it on both classical (ResNet) and modern (ConvNeXt) models: on three downstream tasks, it surpasses both state-of-the-art contrastive learning and transformer-based masked modeling by similarly large margins (around +1.0%). Improvements on object detection and instance segmentation are more substantial (up to +3.5%), verifying the strong transferability of features learned. We also find its favorable scaling behavior by observing more gains on larger models. All this evidence reveals a promising future of generative pre-training on convnets. Codes and models are released at https://github.com/keyu-tian/SparK.

How to use it?

Predict imageUse the modelTrain/Test Command

from mmpretrain import inference_model

predict = inference_model('resnet50_spark-pre_300e_in1k', 'demo/bird.JPEG')
print(predict['pred_class'])
print(predict['pred_score'])

Models and results

Pretrained models

Model	Params (M)	Flops (G)	Config	Download
spark_sparse-resnet50_800e_in1k	37.97	4.10	config	model | log
spark_sparse-convnextv2-tiny_800e_in1k	39.73	4.47	config	model | log

Image Classification on ImageNet-1k

Model	Pretrain	Params (M)	Flops (G)	Top-1 (%)	Top-5 (%)	Config	Download
resnet50_spark-pre_300e_in1k	SPARK	23.52	1.31	80.10	94.90	config	model | log
convnextv2-tiny_spark-pre_300e_in1k	SPARK	28.64	4.47	82.80	96.30	config	model | log

Citation

@Article{tian2023designing,
  author  = {Keyu Tian and Yi Jiang and Qishuai Diao and Chen Lin and Liwei Wang and Zehuan Yuan},
  title   = {Designing BERT for Convolutional Networks: Sparse and Hierarchical Masked Modeling},
  journal = {arXiv:2301.03580},
  year    = {2023},
}