# NeXtVLAD: An Efficient Neural Network to Aggregate Frame-level Features for Large-scale Video Classification

arXiv: https://arxiv.org/abs/1811.05014 · HTML: https://mlsanity.com/paper/1811.05014/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Nov 2018.

## Built on by (most cited first, 24 of 24)

- [NTIRE 2024 Quality Assessment of AI-Generated Content Challenge](https://mlsanity.com/paper/2404.16687/) (2024)
- [Social Fabric: Tubelet Compositions for Video Relation Detection](https://mlsanity.com/paper/2108.08363/) (2021)
- [Temporal-attentive Covariance Pooling Networks for Video Recognition](https://mlsanity.com/paper/2110.14381/) (2021)
- [Alignment-Uniformity aware Representation Learning for Zero-shot Video Classification](https://mlsanity.com/paper/2203.15381/) (2022)
- [Multi-modal Segment Assemblage Network for Ad Video Editing with Importance-Coherence Reward](https://mlsanity.com/paper/2209.12164/) (2022)
- [Alignment-guided Temporal Attention for Video Action Recognition](https://mlsanity.com/paper/2210.00132/) (2022)
- [MM-AU:Towards Multimodal Understanding of Advertisement Videos](https://mlsanity.com/paper/2308.14052/) (2023)
- [Overview of Tencent Multi-modal Ads Video Understanding Challenge](https://mlsanity.com/paper/2109.07951/) (2021)
- [A Survey on Automated Sarcasm Detection on Twitter](https://mlsanity.com/paper/2202.02516/) (2022)
- [Micro-video Tagging via Jointly Modeling Social Influence and Tag Relation](https://mlsanity.com/paper/2303.08318/) (2023)
- [CardioSyntax: end-to-end SYNTAX score prediction -- dataset, benchmark and method](https://mlsanity.com/paper/2407.19894/) (2024)
- [VLAD-BuFF: Burst-aware Fast Feature Aggregation for Visual Place Recognition](https://mlsanity.com/paper/2409.19293/) (2024)
- [Computational Sarcasm Analysis on Social Media: A Systematic Review](https://mlsanity.com/paper/2209.06170/) (2022)
- [Tencent AVS: A Holistic Ads Video Dataset for Multi-modal Scene Segmentation](https://mlsanity.com/paper/2212.04700/) (2022)
- [M3PT: A Multi-Modal Model for POI Tagging](https://mlsanity.com/paper/2306.10079/) (2023)
- [A Multimodal Framework for Video Ads Understanding](https://mlsanity.com/paper/2108.12868/) (2021)
- [Atrial Septal Defect Detection in Children Based on Ultrasound Video Using Multiple Instances Learning](https://mlsanity.com/paper/2306.03835/) (2023)
- [Beyond still images: Temporal features and input variance resilience](https://mlsanity.com/paper/2311.00800/) (2023)
- [Efficient Lung Ultrasound Severity Scoring Using Dedicated Feature Extractor](https://mlsanity.com/paper/2501.12524/) (2025)
- [Deciphering the complaint aspects: Towards an aspect-based complaint identification model with video complaint dataset in finance](https://mlsanity.com/paper/2503.00054/) (2025)
- [UniDot: A Unified Network for Sequence Modeling and Feature Interaction in Large-scale Recommendation](https://mlsanity.com/paper/2608.16797/) (2026)
- [Knowledge Distillation for Efficient Audio-Visual Video Captioning](https://mlsanity.com/paper/2306.09947/) (2023)
- [Enhancing Video Understanding: Deep Neural Networks for Spatiotemporal Analysis](https://mlsanity.com/paper/2502.07277/) (2025)
- [AdaptVPR: Route-Aware Hard Positive Generation for Robust Visual Place Recognition](https://mlsanity.com/paper/2609.04369/) (2026)

## Abstract (arXiv)

This paper introduces a fast and efficient network architecture, NeXtVLAD, to aggregate frame-level features into a compact feature vector for large-scale video classification. Briefly speaking, the basic idea is to decompose a high-dimensional feature into a group of relatively low-dimensional vectors with attention before applying NetVLAD aggregation over time. This NeXtVLAD approach turns out to be both effective and parameter efficient in aggregating temporal information. In the 2nd Youtube-8M video understanding challenge, a single NeXtVLAD model with less than 80M parameters achieves a GAP score of 0.87846 in private leaderboard. A mixture of 3 NeXtVLAD models results in 0.88722, which is ranked 3rd over 394 teams. The code is publicly available at https://github.com/linrongc/youtube-8m.