# Temporal vs. Spatial: Comparing DINOv3 and V-JEPA2 Feature Representations for Video Action Analysis

arXiv: https://arxiv.org/abs/2509.21595 · HTML: https://mlsanity.com/paper/2509.21595/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 39 of its 60 references that are in the map.
- First posted on arXiv 25 Sep 2025.

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

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset](https://mlsanity.com/paper/1705.07750/) (2017)
- [UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction](https://mlsanity.com/paper/1802.03426/) (2018)
- [Exploring Simple Siamese Representation Learning](https://mlsanity.com/paper/2011.10566/) (2020)
- [Non-local Neural Networks](https://mlsanity.com/paper/1711.07971/) (2017)
- [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)
- [ViViT: A Video Vision Transformer](https://mlsanity.com/paper/2103.15691/) (2021)
- [Contrastive Multiview Coding](https://mlsanity.com/paper/1906.05849/) (2019)
- [Curiosity-driven Exploration by Self-supervised Prediction](https://mlsanity.com/paper/1705.05363/) (2017)
- [A Closer Look at Spatiotemporal Convolutions for Action Recognition](https://mlsanity.com/paper/1711.11248/) (2017)
- [Two-Stream Convolutional Networks for Action Recognition in Videos](https://mlsanity.com/paper/1406.2199/) (2014)
- [Deep Variational Information Bottleneck](https://mlsanity.com/paper/1612.00410/) (2016)
- [Temporal Segment Networks: Towards Good Practices for Deep Action Recognition](https://mlsanity.com/paper/1608.00859/) (2016)
- [V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning](https://mlsanity.com/paper/2506.09985/) (2025)
- [Can Spatiotemporal 3D CNNs Retrace the History of 2D CNNs and ImageNet?](https://mlsanity.com/paper/1711.09577/) (2017)
- [Long-term Recurrent Convolutional Networks for Visual Recognition and Description](https://mlsanity.com/paper/1411.4389/) (2014)
- [Look, Listen and Learn](https://mlsanity.com/paper/1705.08168/) (2017)
- [Unsupervised Learning of Video Representations using LSTMs](https://mlsanity.com/paper/1502.04681/) (2015)
- [Generating Videos with Scene Dynamics](https://mlsanity.com/paper/1609.02612/) (2016)
- [Audio-Visual Scene Analysis with Self-Supervised Multisensory Features](https://mlsanity.com/paper/1804.03641/) (2018)
- [FineGym: A Hierarchical Video Dataset for Fine-grained Action Understanding](https://mlsanity.com/paper/2004.06704/) (2020)
- [Beyond Short Snippets: Deep Networks for Video Classification](https://mlsanity.com/paper/1503.08909/) (2015)
- [Self-supervised Co-training for Video Representation Learning](https://mlsanity.com/paper/2010.09709/) (2020)
- [Unsupervised Representation Learning by Sorting Sequences](https://mlsanity.com/paper/1708.01246/) (2017)
- [Self-Supervised Video Representation Learning with Space-Time Cubic Puzzles](https://mlsanity.com/paper/1811.09795/) (2018)

## Built on by (most cited first, 2 of 2)

- [Understanding Multimodal Complementarity for Single-Frame Action Anticipation](https://mlsanity.com/paper/2601.22039/) (2026)
- [Prompt, Probe, Train, or Annotate? Single-camera sports video understanding in amateur settings](https://mlsanity.com/paper/2609.28049/) (2026)

## Abstract (arXiv)

This study presents a comprehensive comparative analysis of two prominent self-supervised learning architectures for video action recognition: DINOv3, which processes frames independently through spatial feature extraction, and V-JEPA2, which employs joint temporal modeling across video sequences. We evaluate both approaches on the UCF Sports dataset, examining feature quality through multiple dimensions including classification accuracy, clustering performance, intra-class consistency, and inter-class discrimination. Our analysis reveals fundamental architectural trade-offs: DINOv3 achieves superior clustering performance (Silhouette score: 0.31 vs 0.21) and demonstrates exceptional discrimination capability (6.16x separation ratio) particularly for pose-identifiable actions, while V-JEPA2 exhibits consistent reliability across all action types with significantly lower performance variance (0.094 vs 0.288). Through action-specific evaluation, we identify that DINOv3's spatial processing architecture excels at static pose recognition but shows degraded performance on motion-dependent actions, whereas V-JEPA2's temporal modeling provides balanced representation quality across diverse action categories. These findings contribute to the understanding of architectural design choices in video analysis systems and provide empirical guidance for selecting appropriate feature extraction methods based on task requirements and reliability constraints.