# Self-Improving 4D Perception via Self-Distillation

arXiv: https://arxiv.org/abs/2604.08532 · HTML: https://mlsanity.com/paper/2604.08532/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Detection and dense prediction distillation, one of 12 directions in Knowledge distillation.
- Builds on 63 of its 79 references that are in the map.
- First posted on arXiv 9 Apr 2026.

## Placed in

- Knowledge distillation / [Detection and dense prediction distillation](https://mlsanity.com/distill/detection-and-dense-prediction-distillation/)

## Builds on (most cited first, 30 of 63)

- [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)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [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)
- [Training data-efficient image transformers & distillation through attention](https://mlsanity.com/paper/2012.12877/) (2020)
- [ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes](https://mlsanity.com/paper/1702.04405/) (2017)
- [CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features](https://mlsanity.com/paper/1905.04899/) (2019)
- [DUSt3R: Geometric 3D Vision Made Easy](https://mlsanity.com/paper/2312.14132/) (2023)
- [VGGT: Visual Geometry Grounded Transformer](https://mlsanity.com/paper/2503.11651/) (2025)
- [SuperGlue: Learning Feature Matching with Graph Neural Networks](https://mlsanity.com/paper/1911.11763/) (2019)
- [Big Self-Supervised Models are Strong Semi-Supervised Learners](https://mlsanity.com/paper/2006.10029/) (2020)
- [Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data](https://mlsanity.com/paper/2401.10891/) (2024)
- [SuperPoint: Self-Supervised Interest Point Detection and Description](https://mlsanity.com/paper/1712.07629/) (2017)
- [Self-training with Noisy Student improves ImageNet classification](https://mlsanity.com/paper/1911.04252/) (2019)
- [Grounding Image Matching in 3D with MASt3R](https://mlsanity.com/paper/2406.09756/) (2024)
- [MVSNet: Depth Inference for Unstructured Multi-view Stereo](https://mlsanity.com/paper/1804.02505/) (2018)
- [Unsupervised Learning of Depth and Ego-Motion from Video](https://mlsanity.com/paper/1704.07813/) (2017)
- [pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesis](https://mlsanity.com/paper/2012.00926/) (2020)
- [Depth Anything 3: Recovering the Visual Space from Any Views](https://mlsanity.com/paper/2511.10647/) (2025)
- [GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis](https://mlsanity.com/paper/2007.02442/) (2020)
- [LightGlue: Local Feature Matching at Light Speed](https://mlsanity.com/paper/2306.13643/) (2023)
- [Continuous 3D Perception Model with Persistent State](https://mlsanity.com/paper/2501.12387/) (2025)
- [Be Your Own Teacher: Improve the Performance of Convolutional Neural Networks via Self Distillation](https://mlsanity.com/paper/1905.08094/) (2019)
- [MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion](https://mlsanity.com/paper/2410.03825/) (2024)
- [Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Pass](https://mlsanity.com/paper/2501.13928/) (2025)
- [HOI4D: A 4D Egocentric Dataset for Category-Level Human-Object Interaction](https://mlsanity.com/paper/2203.01577/) (2022)
- [D2-Net: A Trainable CNN for Joint Detection and Description of Local Features](https://mlsanity.com/paper/1905.03561/) (2019)
- [π3: Permutation-Equivariant Visual Geometry Learning](https://mlsanity.com/paper/2507.13347/) (2025)
- [Learning Category-Specific Mesh Reconstruction from Image Collections](https://mlsanity.com/paper/1803.07549/) (2018)

## Built on by (most cited first, 4 of 4)

- [VGGT-Omega](https://mlsanity.com/paper/2605.15195/) (2026)
- [VLRC: Vision-Language Reprojection Consistency as a scalable signal for better feed-forward 3D pretraining](https://mlsanity.com/paper/2607.02707/) (2026)
- [Self-Geometry: GT-Free and Plug-and-Play Test-Time Adaptation for Geometrically Consistent 3D Vision Foundation Models](https://mlsanity.com/paper/2608.10708/) (2026)
- [Emergent Multi-View Geometry Through Self-Distillation](https://mlsanity.com/paper/2609.39227/) (2026)

## Abstract (arXiv)

Large-scale multi-view reconstruction models have made remarkable progress, but most existing approaches still rely on fully supervised training with ground-truth 3D/4D annotations. Such annotations are expensive and particularly scarce for dynamic scenes, limiting scalability. We propose SelfEvo, a self-improving framework that continually improves pretrained multi-view reconstruction models using unlabeled videos. SelfEvo introduces a self-distillation scheme using spatiotemporal context asymmetry, enabling self-improvement for learning-based 4D perception without external annotations. We systematically study design choices that make self-improvement effective, including loss signals, forms of asymmetry, and other training strategies. Across eight benchmarks spanning diverse datasets and domains, SelfEvo consistently improves pretrained baselines and generalizes across base models (e.g. VGGT and π3), with significant gains on dynamic scenes. Overall, SelfEvo achieves up to 36.5% relative improvement in video depth estimation and 20.1% in camera estimation, without using any labeled data. Project Page: https://self-evo.github.io/.