# The TIME Machine: On The Power of Motion for Efficient Perception

arXiv: https://arxiv.org/abs/2605.23045 · HTML: https://mlsanity.com/paper/2605.23045/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 41 of its 53 references that are in the map.
- First posted on arXiv 21 May 2026.

## Builds on (most cited first, 30 of 41)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [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)
- [Is Space-Time Attention All You Need for Video Understanding?](https://mlsanity.com/paper/2102.05095/) (2021)
- [Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis](https://mlsanity.com/paper/2405.21075/) (2024)
- [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://mlsanity.com/paper/2203.12602/) (2022)
- [Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval](https://mlsanity.com/paper/2104.00650/) (2021)
- [The "something something" video database for learning and evaluating visual common sense](https://mlsanity.com/paper/1706.04261/) (2017)
- [FlowNet: Learning Optical Flow with Convolutional Networks](https://mlsanity.com/paper/1504.06852/) (2015)
- [HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips](https://mlsanity.com/paper/1906.03327/) (2019)
- [V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning](https://mlsanity.com/paper/2506.09985/) (2025)
- [NExT-QA:Next Phase of Question-Answering to Explaining Temporal Actions](https://mlsanity.com/paper/2105.08276/) (2021)
- [Revisiting Feature Prediction for Learning Visual Representations from Video](https://mlsanity.com/paper/2404.08471/) (2024)
- [VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking](https://mlsanity.com/paper/2303.16727/) (2023)
- [CLEVRER: CoLlision Events for Video REpresentation and Reasoning](https://mlsanity.com/paper/1910.01442/) (2019)
- [Masked Autoencoders for Point Cloud Self-supervised Learning](https://mlsanity.com/paper/2203.06604/) (2022)
- [AVA: A Video Dataset of Spatio-temporally Localized Atomic Visual Actions](https://mlsanity.com/paper/1705.08421/) (2017)
- [TempCompass: Do Video LLMs Really Understand Videos?](https://mlsanity.com/paper/2403.00476/) (2024)
- [YouTube-8M: A Large-Scale Video Classification Benchmark](https://mlsanity.com/paper/1609.08675/) (2016)
- [A Short Note on the Kinetics-700 Human Action Dataset](https://mlsanity.com/paper/1907.06987/) (2019)
- [TAPIR: Tracking Any Point with per-frame Initialization and temporal Refinement](https://mlsanity.com/paper/2306.08637/) (2023)
- [YouTube-BoundingBoxes: A Large High-Precision Human-Annotated Data Set for Object Detection in Video](https://mlsanity.com/paper/1702.00824/) (2017)
- [Broaden Your Views for Self-Supervised Video Learning](https://mlsanity.com/paper/2103.16559/) (2021)
- [TemporalBench: Benchmarking Fine-grained Temporal Understanding for Multimodal Video Models](https://mlsanity.com/paper/2410.10818/) (2024)
- [MotionBench: Benchmarking and Improving Fine-grained Video Motion Understanding for Vision Language Models](https://mlsanity.com/paper/2501.02955/) (2025)
- [Only Time Can Tell: Discovering Temporal Data for Temporal Modeling](https://mlsanity.com/paper/1907.08340/) (2019)
- [Self-supervised Video Transformer](https://mlsanity.com/paper/2112.01514/) (2021)
- [VLM4D: Towards Spatiotemporal Awareness in Vision Language Models](https://mlsanity.com/paper/2508.02095/) (2025)
- [Seeing the Arrow of Time in Large Multimodal Models](https://mlsanity.com/paper/2506.03340/) (2025)

## Abstract (arXiv)

Video representation learning has seen tremendous progress in recent years. This has been driven by many factors, including the scale of training and the success of self-supervised models trained on next-frame prediction. While these factors have pushed the boundaries of what video models can do, they also introduce their own set of limitations. First, scaling video models can reach prohibitive costs, with recent models needing hundreds of years of video to be trained. Second, learning to predict the next frame (or embedding of the frame) naturally focuses on spatial information and as a result, video models still struggle with temporal understanding. In this paper we propose a novel approach that uses motion as a modality to alleviate both of these core issues. Given the motion in a video in the form of point tracks, we mask some of the tracks and use a masked autoencoder to reconstruct the missing tracks. This allows us to learn a representation in a self-supervised manner which we call TIME (Temporally-Informed Motion Embedding), and is focused on capturing temporal information. Also, as motion is inherently appearance-invariant, TIME needs far fewer examples to generalize well. As a result, without bells and whistles, on temporal tasks TIME performs on par with state-of-the-art models, using up to 4 orders of magnitude less training data. For general tasks, TIME can be used in combination with existing representations, and we observe that it leads to a significant improvement for V-JEPA 2, RVM and VideoMAE on standard benchmarks such as SSV2, EgoExo4D and Diving48. These results point to a new promising video paradigm for both more temporally-aware as well as more scalable models.