Deep supervised learning algorithms typically require a large volume of labeled data to
achieve satisfactory performance. However, the process of collecting and labeling such data …

The remarkable success of deep learning in various domains relies on the availability of
large-scale annotated datasets. However, obtaining annotations is expensive and requires …

Pre-training video transformers on extra large-scale datasets is generally required to
achieve premier performance on relatively small datasets. In this paper, we show that video …

This paper studies a conceptually simple extension of Masked Autoencoders (MAE) to
spatiotemporal representation learning from videos. We randomly mask out spacetime …

Scale is the primary factor for building a powerful foundation model that could well
generalize to a variety of downstream tasks. However, it is still challenging to train video …

Abstract We present Masked Feature Prediction (MaskFeat) for self-supervised pre-training
of video models. Our approach first randomly masks out a portion of the input sequence and …

Capitalizing on large pre-trained models for various downstream tasks of interest have
recently emerged with promising performance. Due to the ever-growing model size, the …

Video recognition has been dominated by the end-to-end learning paradigm–first initializing
a video recognition model with weights of a pretrained image model and then conducting …

This paper studies the BERT pretraining of video transformers. It is a straightforward but
worth-studying extension given the recent success from BERT pretraining of image …

Establishing correspondence between images or scenes is a significant challenge in
computer vision, especially given occlusions, viewpoint changes, and varying object …