# NMS Threshold matters for Ego4D Moment Queries -- 2nd place solution to the Ego4D Moment Queries Challenge 2023

arXiv: https://arxiv.org/abs/2307.02025 · HTML: https://mlsanity.com/paper/2307.02025/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 11 references that are in the map.
- First posted on arXiv 5 Jul 2023.

## Builds on (most cited first, 10 of 10)

- [FCOS: Fully Convolutional One-Stage Object Detection](https://mlsanity.com/paper/1904.01355/) (2019)
- [YOLOX: Exceeding YOLO Series in 2021](https://mlsanity.com/paper/2107.08430/) (2021)
- [Soft-NMS -- Improving Object Detection With One Line of Code](https://mlsanity.com/paper/1704.04503/) (2017)
- [ActionFormer: Localizing Moments of Actions with Transformers](https://mlsanity.com/paper/2202.07925/) (2022)
- [Egocentric Video-Language Pretraining](https://mlsanity.com/paper/2206.01670/) (2022)
- [Omnivore: A Single Model for Many Visual Modalities](https://mlsanity.com/paper/2201.08377/) (2022)
- [OTA: Optimal Transport Assignment for Object Detection](https://mlsanity.com/paper/2103.14259/) (2021)
- [Diagnosing Error in Temporal Action Detectors](https://mlsanity.com/paper/1807.10706/) (2018)
- [InternVideo-Ego4D: A Pack of Champion Solutions to Ego4D Challenges](https://mlsanity.com/paper/2211.09529/) (2022)
- [Where a Strong Backbone Meets Strong Features -- ActionFormer for Ego4D Moment Queries Challenge](https://mlsanity.com/paper/2211.09074/) (2022)

## Built on by (most cited first, 3 of 3)

- [End-to-End Temporal Action Detection with 1B Parameters Across 1000 Frames](https://mlsanity.com/paper/2311.17241/) (2023)
- [Harnessing Temporal Causality for Advanced Temporal Action Detection](https://mlsanity.com/paper/2407.17792/) (2024)
- [Hier-EgoPack: Hierarchical Egocentric Video Understanding with Diverse Task Perspectives](https://mlsanity.com/paper/2502.02487/) (2025)

## Abstract (arXiv)

This report describes our submission to the Ego4D Moment Queries Challenge 2023. Our submission extends ActionFormer, a latest method for temporal action localization. Our extension combines an improved ground-truth assignment strategy during training and a refined version of SoftNMS at inference time. Our solution is ranked 2nd on the public leaderboard with 26.62% average mAP and 45.69% Recall@1x at tIoU=0.5 on the test set, significantly outperforming the strong baseline from 2023 challenge. Our code is available at https://github.com/happyharrycn/actionformer_release.