# iqiyi Submission to ActivityNet Challenge 2019 Kinetics-700 challenge: Hierarchical Group-wise Attention

Source: [iqiyi Submission to ActivityNet Challenge 2019 Kinetics-700 challenge: Hierarchical Group-wise Attention](<https://www.sota2.com/research/paper/68fa66ff1e209131cea13966>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T22:03:19.943Z

Authors: Qian Liu, Dongyang Cai, Jie Liu, Nan Ding, Tao Wang

Published: 2020-02-07T17:46:38.000Z

## Abstract

In this report, the method for the iqiyi submission to the task of ActivityNet 2019 Kinetics-700 challenge is described. Three models are involved in the model ensemble stage: TSN, HG-NL and StNet. We propose the hierarchical group-wise non-local (HG-NL) module for frame-level features aggregation for video classification. The standard non-local (NL) module is effective in aggregating frame-level features on the task of video classification but presents low parameters efficiency and high computational cost. The HG-NL method involves a hierarchical group-wise structure and generates multiple attention maps to enhance performance. Basing on this hierarchical group-wise structure, the proposed method has competitive accuracy, fewer parameters and smaller computational cost than the standard NL. For the task of ActivityNet 2019 Kinetics-700 challenge, after model ensemble, we finally obtain an averaged top-1 and top-5 error percentage 28.444% on the test set.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2002.02918>)

[PDF](<https://arxiv.org/pdf/2002.02918.pdf>)