# AM Flow: Adapters for Temporal Processing in Action Recognition

arXiv: https://arxiv.org/abs/2411.02065 · HTML: https://mlsanity.com/paper/2411.02065/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 39 of its 52 references that are in the map.
- First posted on arXiv 4 Nov 2024.

## Builds on (most cited first, 30 of 39)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Parameter-Efficient Transfer Learning for NLP](https://mlsanity.com/paper/1902.00751/) (2019)
- [Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset](https://mlsanity.com/paper/1705.07750/) (2017)
- [Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention](https://mlsanity.com/paper/2006.16236/) (2020)
- [SlowFast Networks for Video Recognition](https://mlsanity.com/paper/1812.03982/) (2018)
- [Is Space-Time Attention All You Need for Video Understanding?](https://mlsanity.com/paper/2102.05095/) (2021)
- [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://mlsanity.com/paper/2203.12602/) (2022)
- [Swin Transformer V2: Scaling Up Capacity and Resolution](https://mlsanity.com/paper/2111.09883/) (2021)
- [Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition](https://mlsanity.com/paper/1801.07455/) (2018)
- [The "something something" video database for learning and evaluating visual common sense](https://mlsanity.com/paper/1706.04261/) (2017)
- [Multiscale Vision Transformers](https://mlsanity.com/paper/2104.11227/) (2021)
- [Towards a Unified View of Parameter-Efficient Transfer Learning](https://mlsanity.com/paper/2110.04366/) (2021)
- [TSM: Temporal Shift Module for Efficient Video Understanding](https://mlsanity.com/paper/1811.08383/) (2018)
- [AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognition](https://mlsanity.com/paper/2205.13535/) (2022)
- [X3D: Expanding Architectures for Efficient Video Recognition](https://mlsanity.com/paper/2004.04730/) (2020)
- [Rethinking Spatiotemporal Feature Learning: Speed-Accuracy Trade-offs in Video Classification](https://mlsanity.com/paper/1712.04851/) (2017)
- [MViTv2: Improved Multiscale Vision Transformers for Classification and Detection](https://mlsanity.com/paper/2112.01526/) (2021)
- [Compacter: Efficient Low-Rank Hypercomplex Adapter Layers](https://mlsanity.com/paper/2106.04647/) (2021)
- [Disentangling and Unifying Graph Convolutions for Skeleton-Based Action Recognition](https://mlsanity.com/paper/2003.14111/) (2020)
- [Expanding Language-Image Pretrained Models for General Video Recognition](https://mlsanity.com/paper/2208.02816/) (2022)
- [Prompting Visual-Language Models for Efficient Video Understanding](https://mlsanity.com/paper/2112.04478/) (2021)
- [Hiera: A Hierarchical Vision Transformer without the Bells-and-Whistles](https://mlsanity.com/paper/2306.00989/) (2023)
- [Parameter-efficient Multi-task Fine-tuning for Transformers via Shared Hypernetworks](https://mlsanity.com/paper/2106.04489/) (2021)
- [Temporal Segment Networks for Action Recognition in Videos](https://mlsanity.com/paper/1705.02953/) (2017)
- [UniFormer: Unifying Convolution and Self-attention for Visual Recognition](https://mlsanity.com/paper/2201.09450/) (2022)
- [Frozen CLIP Models are Efficient Video Learners](https://mlsanity.com/paper/2208.03550/) (2022)
- [TAM: Temporal Adaptive Module for Video Recognition](https://mlsanity.com/paper/2005.06803/) (2020)
- [Convolutional Bypasses Are Better Vision Transformer Adapters](https://mlsanity.com/paper/2207.07039/) (2022)
- [Actor-Context-Actor Relation Network for Spatio-Temporal Action Localization](https://mlsanity.com/paper/2006.07976/) (2020)

## Built on by (most cited first, 1 of 1)

- [TALON: Token-Aligned Lightweight Adapters for 6-DoF Spacecraft Pose Estimation](https://mlsanity.com/paper/2605.31217/) (2026)

## Abstract (arXiv)

Deep learning models, in particular textitimage models, have recently gained generalisability and robustness. %are becoming more general and robust by the day. In this work, we propose to exploit such advances in the realm of textitvideo classification. Video foundation models suffer from the requirement of extensive pretraining and a large training time. Towards mitigating such limitations, we propose "textitAttention Map (AM) Flow" for image models, a method for identifying pixels relevant to motion in each input video frame. In this context, we propose two methods to compute AM flow, depending on camera motion. AM flow allows the separation of spatial and temporal processing, while providing improved results over combined spatio-temporal processing (as in video models). Adapters, one of the popular techniques in parameter efficient transfer learning, facilitate the incorporation of AM flow into pretrained image models, mitigating the need for full-finetuning. We extend adapters to "textittemporal processing adapters" by incorporating a temporal processing unit into the adapters. Our work achieves faster convergence, therefore reducing the number of epochs needed for training. Moreover, we endow an image model with the ability to achieve state-of-the-art results on popular action recognition datasets. This reduces training time and simplifies pretraining. We present experiments on Kinetics-400, Something-Something v2, and Toyota Smarthome datasets, showcasing state-of-the-art or comparable results.