Memento: Reconstruct to Remember for Consistent Long Video Generation
Summary
Memento is a subject-reconstruction-guided framework that improves long-form video generation by preserving recurring subjects through memory-based reconstruction and dual-query mechanisms, achieving state-of-the-art performance in long-term subject consistency and cross-shot coherence.
View Cached Full Text
Cached at: 06/16/26, 11:31 AM
Paper page - Memento: Reconstruct to Remember for Consistent Long Video Generation
Source: https://huggingface.co/papers/2606.14667
Abstract
Memento is a subject-reconstruction-guided framework that improves long-form video generation by preserving recurring subjects through memory-based reconstruction and dual-query mechanisms.
Long-form video generation requires recurring subjects to remain consistent across various shots, viewpoints, motions, and scene transitions. Existingtemporal decompositionmethods improve scalability by generating videos shot by shot. However, they mainly focus on optimizing plausible next-shot continuations without verifying whether the historical memory preserves identity-critical subject evidence. Consequently, as generation proceeds, recurring subjects may be diluted, overwritten, or forgotten. In this paper, we propose Memento, a subject-reconstruction-guided framework that treats subject preservation as an explicit identity grounding problem, based on the premise that amemory bankfaithfully preserving a subject should support reconstructing that subject from memory alone. Specifically, Memento jointly trainsautoregressive next-shot generationwith memory-basedsubject reconstruction, recovering target appearances using historical memory and global story captions. To disentangle long-range subject evidence from short-range cues, Memento introduces adual-query memory mechanism, where one query retrieves identity-relevant memory and the other selects short-context keyframes for coherent continuation. Additionally, a subject-awarecinematic data pipelineprovides precise reconstruction supervision via consistent, pronoun-free subject descriptions. Experiments demonstrate that Memento achieves state-of-the-art performance inlong-term subject consistency,cross-shot coherence, andvisual quality.
View arXiv pageView PDFProject pageGitHub1Add to collection
Get this paper in your agent:
hf papers read 2606\.14667
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper1
#### ernie-research/Memento Text-to-Video• Updatedabout 2 hours ago • 19 • 3
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2606.14667 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2606.14667 in a Space README.md to link it from this page.
Collections including this paper2
Similar Articles
MemLife: Curating and Reasoning over Long-Term Egocentric Video Memories
MemLife introduces a multimodal memory system for long-term egocentric videos that builds entity-grounded text episodes and retrieves them with a time-indexed agentic reader, improving over training-free baselines by 4.6-12.0% on long-horizon benchmarks. A reinforcement learning framework called MemOpt further optimizes the memory writer for faithfulness and retrievability, yielding consistent 2.7-5.0% gains.
Reasoning with Memory: A Temporal Granularity-Adaptive Framework for Training-Free Long Video Understanding
ReMem introduces a dual-level memory-augmented keyframe selection framework for training-free long video understanding, achieving state-of-the-art zero-shot performance on multiple benchmarks.
PermaVid: Consistent Video Generation Across Edits via Disentangled Context Memory
PermaVid introduces a multi-modal context memory that disentangles appearance and geometric structure to maintain long-term video consistency after editing operations, outperforming prior methods.
The Past Frames the Future: Memory for Autoregressive Video Generation
This paper provides a comprehensive review of memory mechanisms for autoregressive video generation, addressing the challenge of maintaining historical information across extended sequences to ensure temporal persistence.
DecMem: Towards Minute-Long Consistent World Generation with Decoupled Memory
DecMem introduces a decoupled memory architecture with Sparse Global Memory and Anchored Local Memory to achieve consistent minute-long video generation, outperforming state-of-the-art methods.