# CascadeFormer: A Family of Two-stage Cascading Transformers for Skeleton-based Human Action Recognition

arXiv: https://arxiv.org/abs/2509.00692 · HTML: https://mlsanity.com/paper/2509.00692/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 19 of its 29 references that are in the map.
- First posted on arXiv 31 Aug 2025.

## Builds on (most cited first, 19 of 19)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Is Space-Time Attention All You Need for Video Understanding?](https://mlsanity.com/paper/2102.05095/) (2021)
- [Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition](https://mlsanity.com/paper/1801.07455/) (2018)
- [An overview of gradient descent optimization algorithms](https://mlsanity.com/paper/1609.04747/) (2016)
- [DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification](https://mlsanity.com/paper/2106.02034/) (2021)
- [NTU RGB+D: A Large Scale Dataset for 3D Human Activity Analysis](https://mlsanity.com/paper/1604.02808/) (2016)
- [ImageNet-21K Pretraining for the Masses](https://mlsanity.com/paper/2104.10972/) (2021)
- [Channel-wise Topology Refinement Graph Convolution for Skeleton-Based Action Recognition](https://mlsanity.com/paper/2107.12213/) (2021)
- [Spatio-Temporal LSTM with Trust Gates for 3D Human Action Recognition](https://mlsanity.com/paper/1607.07043/) (2016)
- [Cross-view Action Modeling, Learning and Recognition](https://mlsanity.com/paper/1405.2941/) (2014)
- [Temporal Query Networks for Fine-grained Video Understanding](https://mlsanity.com/paper/2104.09496/) (2021)
- [OmniVec: Learning robust representations with cross modal sharing](https://mlsanity.com/paper/2311.05709/) (2023)
- [Cross-Modal Learning with 3D Deformable Attention for Action Recognition](https://mlsanity.com/paper/2212.05638/) (2022)
- [OmniVec2 -- A Novel Transformer based Network for Large Scale Multimodal and Multitask Learning](https://mlsanity.com/paper/2507.13364/) (2025)
- [Ske2Grid: Skeleton-to-Grid Representation Learning for Action Recognition](https://mlsanity.com/paper/2308.07571/) (2023)

## Abstract (arXiv)

Skeleton-based human action recognition leverages sequences of human joint coordinates to identify actions performed in videos. Owing to the intrinsic spatiotemporal structure of skeleton data, Graph Convolutional Networks (GCNs) have been the dominant architecture in this field. However, recent advances in transformer models and masked pretraining frameworks open new avenues for representation learning. In this work, we propose CascadeFormer, a family of two-stage cascading transformers for skeleton-based human action recognition. Our framework consists of a masked pretraining stage to learn generalizable skeleton representations, followed by a cascading fine-tuning stage tailored for discriminative action classification. We evaluate CascadeFormer across three benchmark datasets (Penn Action N-UCLA, and NTU RGB+D 60), achieving competitive performance on all tasks. To promote reproducibility, we release our code and model checkpoints.