Temporal Aggregate Representations for Long-Range Video Understanding
In proceedings of European Conference on Computer Vision (ECCV), 2020
Abstract
Future prediction, especially in long-range videos, requires reasoning from current and past observations. In this work, we address questions of temporal extent, scaling, and level of semantic abstraction with a flexible multi-granular temporal aggregation framework. We show that it is possible to achieve state of the art in both next action and dense anticipation with simple techniques such as max-pooling and attention. To demonstrate the anticipation capabilities of our model, we conduct experiments on Breakfast, 50Salads, and EPIC-Kitchens datasets, where we achieve state-of-the-art results. With minimal modifications, our model can also be extended for video segmentation and action recognition.
Images
![]() |
Download Paper
Additional Material
- Software (ZIP archive, 25 KB)
Bibtex
@INPROCEEDINGS{sener-2020-temporal, author = {Sener, Fadime and Singhania, Dipika and Yao, Angela}, title = {Temporal Aggregate Representations for Long-Range Video Understanding}, booktitle = {European Conference on Computer Vision (ECCV)}, year = {2020}, abstract = {Future prediction, especially in long-range videos, requires reasoning from current and past observations. In this work, we address questions of temporal extent, scaling, and level of semantic abstraction with a flexible multi-granular temporal aggregation framework. We show that it is possible to achieve state of the art in both next action and dense anticipation with simple techniques such as max-pooling and attention. To demonstrate the anticipation capabilities of our model, we conduct experiments on Breakfast, 50Salads, and EPIC-Kitchens datasets, where we achieve state-of-the-art results. With minimal modifications, our model can also be extended for video segmentation and action recognition.} }