# Temporal Action Localization in Untrimmed Videos via Multi-stage CNNs

arXiv: https://arxiv.org/abs/1601.02129 · HTML: https://mlsanity.com/paper/1601.02129/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 9 Jan 2016.

## Built on by (most cited first, 5 of 5)

- [Read and Attend: Temporal Localisation in Sign Language Videos](https://mlsanity.com/paper/2103.16481/) (2021)
- [Fine-grained Temporal Contrastive Learning for Weakly-supervised Temporal Action Localization](https://mlsanity.com/paper/2203.16800/) (2022)
- [Hear Me Out: Fusional Approaches for Audio Augmented Temporal Action Localization](https://mlsanity.com/paper/2106.14118/) (2021)
- [LocATe: End-to-end Localization of Actions in 3D with Transformers](https://mlsanity.com/paper/2203.10719/) (2022)
- [Few-Shot Transformation of Common Actions into Time and Space](https://mlsanity.com/paper/2104.02439/) (2021)

## Abstract (arXiv)

We address temporal action localization in untrimmed long videos. This is important because videos in real applications are usually unconstrained and contain multiple action instances plus video content of background scenes or other activities. To address this challenging issue, we exploit the effectiveness of deep networks in temporal action localization via three segment-based 3D ConvNets: (1) a proposal network identifies candidate segments in a long video that may contain actions; (2) a classification network learns one-vs-all action classification model to serve as initialization for the localization network; and (3) a localization network fine-tunes on the learned classification network to localize each action instance. We propose a novel loss function for the localization network to explicitly consider temporal overlap and therefore achieve high temporal localization accuracy. Only the proposal network and the localization network are used during prediction. On two large-scale benchmarks, our approach achieves significantly superior performances compared with other state-of-the-art systems: mAP increases from 1.7% to 7.4% on MEXaction2 and increases from 15.0% to 19.0% on THUMOS 2014, when the overlap threshold for evaluation is set to 0.5.