UnLoc: A Unified Framework for Video Localization Tasks

TL;DR

UnLoc introduces a unified, single-stage framework leveraging pretrained image and text models for multiple video localization tasks, achieving state-of-the-art results without task-specific components.

Contribution

It is the first unified model capable of Moment Retrieval, Temporal Localization, and Action Segmentation in untrimmed videos using a single architecture.

Findings

State-of-the-art performance on all three localization tasks

No need for action proposals or motion-based features

Unified approach simplifies video localization pipeline

Abstract

While large-scale image-text pretrained models such as CLIP have been used for multiple video-level tasks on trimmed videos, their use for temporal localization in untrimmed videos is still a relatively unexplored task. We design a new approach for this called UnLoc, which uses pretrained image and text towers, and feeds tokens to a video-text fusion model. The output of the fusion module are then used to construct a feature pyramid in which each level connects to a head to predict a per-frame relevancy score and start/end time displacements. Unlike previous works, our architecture enables Moment Retrieval, Temporal Localization, and Action Segmentation with a single stage model, without the need for action proposals, motion based pretrained features or representation masking. Unlike specialized models, we achieve state of the art results on all three different localization tasks with a unified approach. Code will be available at: \url{https://github.com/google-research/scenic}.