# PRISM: Predictive Recomposition via Semantic Latent Decomposition for View-invariant Video Representation Learning

arXiv: https://arxiv.org/abs/2608.30388 · HTML: https://mlsanity.com/paper/2608.30388/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 17 of its 27 references that are in the map.
- First posted on arXiv 31 Aug 2026.

## Builds on (most cited first, 17 of 17)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Is Space-Time Attention All You Need for Video Understanding?](https://mlsanity.com/paper/2102.05095/) (2021)
- [Learning Important Features Through Propagating Activation Differences](https://mlsanity.com/paper/1704.02685/) (2017)
- [SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features](https://mlsanity.com/paper/2502.14786/) (2025)
- [Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives](https://mlsanity.com/paper/2311.18259/) (2023)
- [Time-Contrastive Networks: Self-Supervised Learning from Video](https://mlsanity.com/paper/1704.06888/) (2017)
- [Assembly101: A Large-Scale Multi-View Video Dataset for Understanding Procedural Activities](https://mlsanity.com/paper/2203.14712/) (2022)
- [Temporal Cycle-Consistency Learning](https://mlsanity.com/paper/1904.07846/) (2019)
- [Charades-Ego: A Large-Scale Dataset of Paired Third and First Person Videos](https://mlsanity.com/paper/1804.09626/) (2018)
- [Actor and Observer: Joint Modeling of First and Third-Person Videos](https://mlsanity.com/paper/1804.09627/) (2018)
- [Learning Fine-grained View-Invariant Representations from Unpaired Ego-Exo Videos via Temporal Alignment](https://mlsanity.com/paper/2306.05526/) (2023)
- [Representation Learning via Global Temporal Alignment and Cycle-Consistency](https://mlsanity.com/paper/2105.05217/) (2021)
- [Frame-wise Action Representations for Long Videos via Sequence Contrastive Learning](https://mlsanity.com/paper/2203.14957/) (2022)
- [MASH-VLM: Mitigating Action-Scene Hallucination in Video-LLMs through Disentangled Spatial-Temporal Representations](https://mlsanity.com/paper/2503.15871/) (2025)
- [Learning from Semantic Alignment between Unpaired Multiviews for Egocentric Video Recognition](https://mlsanity.com/paper/2308.11489/) (2023)
- [DEVIAS: Learning Disentangled Video Representations of Action and Scene](https://mlsanity.com/paper/2312.00826/) (2023)

## Abstract (arXiv)

Cross-view video representation learning aims to capture viewpoint-invariant action semantics despite substantial appearance changes across egocentric and exocentric videos. However, existing methods encode each video as a unified embedding, where view-invariant and view-variant semantics inevitably entangle under co-occurrences - a failure mode we show persists even in cross-view methods explicitly trained for view-invariance. Our key insight is that a view-invariant feature is truly disentangled when it can be sufficiently recomposed with an arbitrary view-variant feature while preserving their independent semantics. Building on this, we propose PRISM, that decomposes video into view-invariant and view-variant latents and recompose them under language supervision encouraging clean decomposition of the two streams. PRISM achieves state-of-the-art results on EgoExo4D, EgoExoLearn, AE2, even surpassing in-domain models under zero-shot setting. Code is available at https://github.com/litcoderr/prism.