# Knowledge Integration Networks for Action Recognition

arXiv: https://arxiv.org/abs/2002.07471 · HTML: https://mlsanity.com/paper/2002.07471/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Feb 2020.

## Built on by (most cited first, 8 of 8)

- [Cross-Modal Knowledge Distillation Method for Automatic Cued Speech Recognition](https://mlsanity.com/paper/2106.13686/) (2021)
- [CDST: Color Disentangled Style Transfer for Universal Style Reference Customization](https://mlsanity.com/paper/2506.13770/) (2025)
- [Hyper-parameter tuning for text guided image editing](https://mlsanity.com/paper/2407.21703/) (2024)
- [NAS-TC: Neural Architecture Search on Temporal Convolutions for Complex Action Recognition](https://mlsanity.com/paper/2104.01110/) (2021)
- [Partial to Whole Knowledge Distillation: Progressive Distilling Decomposed Knowledge Boosts Student Better](https://mlsanity.com/paper/2109.12507/) (2021)
- [Practical Knowledge Distillation: Using DNNs to Beat DNNs](https://mlsanity.com/paper/2302.12360/) (2023)
- [Style-CCL: Content-Preserving Style Transfer via Curriculum Continual Learning](https://mlsanity.com/paper/2606.14746/) (2026)
- [TeleStyle V2: Beyond Content-Preserving Style Transfer with Self-Distillation and Distribution-Matching-Distillation](https://mlsanity.com/paper/2606.20709/) (2026)

## Abstract (arXiv)

In this work, we propose Knowledge Integration Networks (referred as KINet) for video action recognition. KINet is capable of aggregating meaningful context features which are of great importance to identifying an action, such as human information and scene context. We design a three-branch architecture consisting of a main branch for action recognition, and two auxiliary branches for human parsing and scene recognition which allow the model to encode the knowledge of human and scene for action recognition. We explore two pre-trained models as teacher networks to distill the knowledge of human and scene for training the auxiliary tasks of KINet. Furthermore, we propose a two-level knowledge encoding mechanism which contains a Cross Branch Integration (CBI) module for encoding the auxiliary knowledge into medium-level convolutional features, and an Action Knowledge Graph (AKG) for effectively fusing high-level context information. This results in an end-to-end trainable framework where the three tasks can be trained collaboratively, allowing the model to compute strong context knowledge efficiently. The proposed KINet achieves the state-of-the-art performance on a large-scale action recognition benchmark Kinetics-400, with a top-1 accuracy of 77.8%. We further demonstrate that our KINet has strong capability by transferring the Kinetics-trained model to UCF-101, where it obtains 97.8% top-1 accuracy.