# Cycle Consistency in Video Object-Centric Learning

arXiv: https://arxiv.org/abs/2605.30211 · HTML: https://mlsanity.com/paper/2605.30211/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 27 references that are in the map.
- First posted on arXiv 28 May 2026.

## Builds on (most cited first, 25 of 25)

- [Object-Centric Learning with Slot Attention](https://mlsanity.com/paper/2006.15055/) (2020)
- [Conditional Object-Centric Learning from Video](https://mlsanity.com/paper/2111.12594/) (2021)
- [Bridging the Gap to Real-World Object-Centric Learning](https://mlsanity.com/paper/2209.14860/) (2022)
- [SAVi++: Towards End-to-End Object-Centric Learning from Real-World Videos](https://mlsanity.com/paper/2206.07764/) (2022)
- [Illiterate DALL-E Learns to Compose](https://mlsanity.com/paper/2110.11405/) (2021)
- [SlotFormer: Unsupervised Visual Dynamics Simulation with Object-Centric Models](https://mlsanity.com/paper/2210.05861/) (2022)
- [Simple Unsupervised Object-Centric Learning for Complex and Naturalistic Videos](https://mlsanity.com/paper/2205.14065/) (2022)
- [SlotDiffusion: Object-Centric Generative Modeling with Diffusion Models](https://mlsanity.com/paper/2305.11281/) (2023)
- [Object-Centric Learning for Real-World Videos by Predicting Temporal Feature Similarities](https://mlsanity.com/paper/2306.04829/) (2023)
- [Improving Object-centric Learning with Query Optimization](https://mlsanity.com/paper/2210.08990/) (2022)
- [Invariant Slot Attention: Object Discovery with Slot-Centric Reference Frames](https://mlsanity.com/paper/2302.04973/) (2023)
- [SPOT: Self-Training with Patch-Order Permutation for Object-Centric Learning with Autoregressive Transformers](https://mlsanity.com/paper/2312.00648/) (2023)
- [Provable Compositional Generalization for Object-Centric Learning](https://mlsanity.com/paper/2310.05327/) (2023)
- [Adaptive Slot Attention: Object Discovery with Dynamic Slot Number](https://mlsanity.com/paper/2406.09196/) (2024)
- [Temporally Consistent Object-Centric Learning by Contrasting Slots](https://mlsanity.com/paper/2412.14295/) (2024)
- [Self-Supervised Multi-Object Tracking with Cross-Input Consistency](https://mlsanity.com/paper/2111.05943/) (2021)
- [Self-Supervised Multi-Object Tracking with Path Consistency](https://mlsanity.com/paper/2404.05136/) (2024)
- [Object-Centric Multiple Object Tracking](https://mlsanity.com/paper/2309.00233/) (2023)
- [Slot Attention with Re-Initialization and Self-Distillation](https://mlsanity.com/paper/2507.23755/) (2025)
- [Predicting Video Slot Attention Queries from Random Slot-Feature Pairs](https://mlsanity.com/paper/2508.01345/) (2025)
- [Vector-Quantized Vision Foundation Models for Object-Centric Learning](https://mlsanity.com/paper/2502.20263/) (2025)
- [Tracking without Label: Unsupervised Multiple Object Tracking via Contrastive Similarity Learning](https://mlsanity.com/paper/2309.00942/) (2023)
- [Smoothing Slot Attention Iterations and Recurrences](https://mlsanity.com/paper/2508.05417/) (2025)
- [Walker: Self-supervised Multiple Object Tracking by Walking on Temporal Appearance Graphs](https://mlsanity.com/paper/2409.17221/) (2024)
- [YOLO11-JDE: Fast and Accurate Multi-Object Tracking with Self-Supervised Re-ID](https://mlsanity.com/paper/2501.13710/) (2025)

## Abstract (arXiv)

Self-supervised video Object-Centric Learning (OCL) aims to discover distinct objects and associate them across time, whereas self-supervised Multi-Object Tracking (MOT) focuses on associating pre-defined object detections or segmentations. Although well-established in MOT, Cycle Consistency (CC) cannot naively or explicitly apply to the latent slot space of OCL. Unlike the deterministic and ideal object representations in MOT, OCL slots are inherently stochastic and ambiguous due to non-unique scene decompositions. Enforcing explicit cycle consistency (ECC) on slots imposes rigid mean seeking. This severely penalizes the model for exploring alternative but equally valid decompositions, thereby driving towards feature collapse. To resolve this dilemma, we propose textitImplicit Cycle Consistency (ICC), which shifts the cycle-consistency constraint from the restrictive slot space to the continuous reconstruction manifold, encouraging slots to reach a soft consensus on collectively interpreting the visual scene rather than forcing rigid point-to-point feature alignment. Extensive experiments on complex video OCL benchmarks demonstrate that ICC avoids feature collapse and outperforms ECC baselines. Our source code, model checkpoints and training logs are provided on https://github.com/Genera1Z/ICC.