Skip to yearly menu bar Skip to main content


Rethinking Video ViTs: Sparse Video Tubes for Joint Image and Video Learning

AJ Piergiovanni · Weicheng Kuo · Anelia Angelova

West Building Exhibit Halls ABC 212


We present a simple approach which can turn a ViT encoder into an efficient video model, which can seamlessly work with both image and video inputs. By sparsely sampling the inputs, the model is able to do training and inference from both inputs. The model is easily scalable and can be adapted to large-scale pre-trained ViTs without requiring full finetuning. The model achieves SOTA results.

Chat is not available.